Information processing methods

Abstract

The multiple asset elements constituting an intangible asset and their combined state are understood as a dynamic state including changes over time, and future states can be explored as scenarios with multiple branches. [Solution] For multiple asset elements 201 to 205 constituting the intangible asset 200, an index 207 is calculated based on the combined state 206 and placed in a multidimensional latent space 210. The index 207 placed in the latent space 210 is converted to an observation space 220 based on the user's instructions and displayed visually. An index corresponding to a future state is generated based on a state transition model and presented as a future scenario 300 including multiple state transition branches 301. Future state generation can be performed using an externally provided calculation module 40.

Description

Technical Field

[0001] The present invention relates to intangible assets owned or managed by an individual, corporation, organization, or group, calculates indices based on a plurality of asset elements constituting the intangible assets and their combined states, and arranges the indices in a multi-dimensional latent space to enable exploration of temporal changes and future state bifurcations, and relates to an information processing method therefor.

Background Art

[0002] In recent years, in corporate activities and various organizational operations, in addition to tangible assets such as facilities and funds, intangible assets such as know-how, technical ideas, organizational capabilities, human networks, customer relationships, and brand values are being recognized as elements that form the core of competitive advantage and value creation.

[0003] These intangible assets often exhibit effective value only when a plurality of asset elements are combined with each other and function under specific market environments or subject conditions, rather than existing as a single element.

[0004] In addition, legal rights such as patent rights, utility model rights, design rights, and trademark rights, as well as mechanisms such as contracts, trade secret management systems, and brand management, are positioned as means for protecting the value of intangible assets or controlling access and use by third parties. Although these protection means have different natures from the intangible assets themselves, they can function as elements that directly affect the market value, competitive position, and risk structure of intangible assets.

[0005] (Prior Art Regarding Latent Space and Observation Space) In the field of machine learning, a method has been proposed for encoding high-dimensional data into a low-dimensional latent space and predicting future states in the latent space (see Patent Document 1 and Non-Patent Document 1). In addition, in a state space model, a structure has been formulated in which a part of the state space is selectively observed by an observation matrix (selection matrix) that defines a mapping from a state vector to an observation vector (see Non-Patent Document 2).

[0006] However, many of these prior arts deal with data of a single modality, such as images, audio, and natural language. Technologies for comprehensively representing intangible assets—where multiple heterogeneous asset elements (legal rights, know-how, brand value, etc.) combine to form value—in latent space are not yet fully established. Furthermore, technical considerations from the perspective of presenting the state in a format that users can intuitively understand during the conversion from latent space to observed space are limited.

[0007] (Prior art concerning the combined state of multiple elements) In the field of corporate or organizational evaluation, a method has been proposed in which potential factors are extracted from multiple evaluation indicators through factor analysis, the correspondence between these factors and the evaluation indicators is calculated as factor loadings, and an overall evaluation indicator is calculated using the extracted factors (see Patent Document 2). In this method, the correspondence between evaluation indicators and factors is not limited to one-to-one relationships, but is treated as a many-to-many relationship where multiple evaluation indicators contribute to a single factor, or a single evaluation indicator contributes to multiple factors.

[0008] Furthermore, in the field of decision support, a method has been proposed that estimates a multi-attribute utility function from the preference structure for multiple attributes and aggregates the multiple attributes into a single utility value using this utility function (see Patent Document 3). This method provides a technical foundation for quantifying the combined state of multiple elements in a form that can be calculated by weighted aggregation operations.

[0009] However, much of this prior art deals with static evaluations at a specific point in time, and there is a lack of established technology to continuously track and display the dynamic structure in which the combined state of multiple asset elements changes over time, in response to changes in the external environment, or in response to the actions of the entities. In particular, with intangible assets, the combined state can change discontinuously due to the expiration of legal rights, obsolescence of technology, changes in the market environment, the entry of competitors, etc., but there are limited technologies that can handle dynamic structures including such discontinuous changes.

[0010] (Prior art relating to future state prediction and branching scenarios) In the field of time series analysis, numerous methods have been proposed to predict future states based on past observational data. Furthermore, in the field of scenario planning, methods are known for qualitatively describing multiple future scenarios and using them as a basis for strategic planning.

[0011] However, when predicting the future state of intangible assets, numerous external factors such as market conditions, competitive landscape, technological trends, and legal regulations have a complex interplay of influences. Therefore, it is desirable to present future scenarios that include multiple branches rather than a single predicted value. In prior art, there is no well-established technology that visually presents such future scenarios including multiple branches along with their probability of occurrence, and that allows users to explore the details of each branch.

[0012] Furthermore, when generating future states, there are limited technologies that allow for the setting of branching conditions based on similarities with past cases (including success and failure cases) and the explanation of the reason for the generation of such a branch to the user.

[0013] (Prior technologies related to external module integration) In the field of decision management, there is a known technology that integrates decision-making modules such as predictive models, optimization algorithms, business rules, decision tables, and scorecards as services on a platform, and structures these modules using industry-standard formats such as PMML (Predictive Model Markup Language), XML, and JSON (see Patent Document 4). Such technology provides a foundation that allows externally provided computation modules to be called without disclosing their internal implementation.

[0014] Furthermore, in cloud-based decision-making platforms, technologies have been proposed to add third-party decision-making modules to the platform and integrate those modules into analytics-driven applications.

[0015] However, these prior arts primarily focus on routine decision-making processes such as credit assessment, fraud detection, and marketing optimization. Their application to non-routine valuations, such as the valuation of intangible assets, which require consideration of diverse factors including the combined state of multiple asset elements, the attributes of the valuation body, and the market environment, has not been adequately explored. In particular, the configuration of integrating the output obtained from external modules as indicators in a latent space and using them to generate future states is not explicitly stated in the prior art.

[0016] (Prior art regarding comparison of evaluation results and uncertainty) In the field of metrology, methods have been established to ensure the reliability of measurement results by evaluating the uncertainty of the measurement results and clearly specifying the measurement resolution and measurement range. Furthermore, in the field of statistics, methods for quantitatively expressing the uncertainty of estimates by assigning confidence intervals to estimated values ​​are widely used.

[0017] However, in the valuation of intangible assets, the accuracy and reliability of the valuation results can vary greatly depending on the nature of the asset being valued, the choice of valuation method, and the attributes of the valuation body. In the prior art, techniques for explicitly indicating such uncertainty in valuation results using metrological concepts such as measurement resolution, measurement range, and error band, and for treating differences below the measurement resolution as indistinguishable (e.g., identical representation, equiranking), are not yet well established in the field of intangible asset valuation.

[0018] In particular, when comparing multiple future scenarios, the technology for determining the significance of differences based on branching points and highlighting only differences exceeding the measurement resolution as statistically significant is not explicitly stated in prior art.

[0019] (Prior art relating to the acquisition and re-evaluation of external information) In event-driven information processing systems, technologies that detect external events (triggers) and execute predetermined processes in response to those events are widely used. Furthermore, in the field of real-time analysis, technologies that continuously monitor changes in external data and update analysis results when changes are detected are known.

[0020] However, in the valuation of intangible assets, the degree to which changes in the external environment (such as legal and regulatory changes, competitor activity, technological trends, and market environment changes) influence the valuation results (importance) can vary depending on the nature of the asset being valued, the attributes of the valuation body, and the valuation objectives. In prior art, there is no well-established technology that calculates the importance of external trigger information, updates the indicators and regenerates future scenarios only when the importance meets predetermined conditions, and visually presents the difference before and after the update.

[0021] (Prior technology related to reliability indication using facial expressions) In the field of conversational agents, there is a known technique that influences the trustworthiness or credibility perceived by users by switching the facial expressions (demeanors) of avatars (see Patent Document 5). This technique suggests that avatars with specific facial expressions, such as smiles, can gain higher trust from users compared to avatars with neutral facial expressions.

[0022] Furthermore, there is a known technique that generates facial expression parameters, including eye movements, mouth movements, and head posture, using text, audio, or video as input, and renders an avatar based on these parameters (see Patent Document 6). This technique extracts angle information that constitutes facial expressions from the landmark coordinates of the face and makes it possible to apply it to two-dimensional or three-dimensional character representations.

[0023] However, these prior arts mainly aim at interactive agents or communication support, and in the evaluation of intangible assets, visualization of audit value, or inter-subject credit risk assessment, they are not well-established enough as technologies to express the degree of deviation from the criteria or the application status of double criteria as expressions and facilitate intuitive understanding of users.

[0024] (Summary of Prior Arts) As described above, intangible assets, their combination states, protection means, and the market environment and external conditions surrounding them change dynamically in response to the passage of time, actions of entities, institutional changes, technological progress, etc. Therefore, it is difficult to fully understand future value fluctuations, risks, and opportunities by only grasping the state of intangible assets as static lists or single indicators.

[0025] Conventionally, as methods for grasping and analyzing intangible assets, patent maps, IP landscapes, technology roadmaps, financial indicator analyses, market analysis reports, etc. have been widely used. These methods have certain usefulness in visualizing the distribution of intangible assets and market conditions at a specific point in time.

[0026] However, many of the prior arts have limitations in the following aspects. First, it is difficult to integrally handle the combination state between multiple elements constituting intangible assets, the relationship with protection means, and the relativity of value due to evaluation entities and market conditions within the same framework. Second, they do not fully have the function of systematically exploring and comparing state transitions along the time axis and multiple branch scenarios in the future. Third, although future prediction technologies using machine learning models or generative models have been proposed, their internal structures are likely to be black-boxed, and it is difficult for users to grasp the influence of differences in evaluation criteria and preconditions on the results. Fourth, technologies for integrally using externally provided computing modules in the evaluation of intangible assets and the generation of future states are not well-established enough. Fifthly, in the comparison of evaluation results, techniques for applying metrological concepts such as measurement resolution, measurement range, error band, etc. to improve the robustness, reproducibility, and accountability of judgments have not been fully established. Sixthly, techniques for dynamically updating evaluation results according to changes in the external environment and visually presenting the differences before and after the update to assist users in making decisions have not been fully established.

[0027] In particular, techniques for explicitly handling fluctuations in evaluation results due to differences in the attributes and normative criteria of evaluation subjects and providing them in a form that users can intuitively operate and understand have not been fully established.

Prior Art Documents

Patent Documents

[0028]

Patent Document 1

Patent Document 2

Patent Document 3

Patent Document 4

Patent Document 5

Patent Document 6

Non-Patent Documents

[0029]

Non-Patent Document 1

Non-Patent Document 2

[0030] The present invention aims to solve at least some or all of the following problems. (a) It is difficult to grasp the multiple asset elements that constitute an intangible asset, their combined state, and the relationship between protective measures and the market environment, not as a static representation at a single point in time, but as a dynamic state that includes changes over time. (b) Despite the fact that evaluation results for the same subject may vary due to differences in the attributes of the evaluator, normative standards, or purpose of use, it is difficult to explicitly address the impact of such differences on the results and present them in a way that users can intuitively understand and manipulate. (c) It is difficult to explore and compare future states as multiple branching and uncertain scenarios. (d) The method for calculating the indicators, the state transition model, or the evaluation model tend to become fixed, making it difficult to flexibly replace or adjust them according to differences in application, purpose, and evaluation perspective. (e) The evaluation results of distance or similarity used for comparison or ranking may be subject to ambiguity regarding the handling of measurement resolution, measurement range, or error, which may reduce the robustness of judgment, accountability, or reproducibility.

[0031] The present invention aims to solve these problems and provide information processing technology that makes intangible assets, protective measures, and related markets intuitively understandable as a dynamic state including evaluation entities and time axes, enabling the exploration and comparison of future states. [Means for solving the problem]

[0032] The information processing method according to the present invention acquires information on multiple asset elements constituting an intangible asset and their combined state, as well as rights or protective measures associated with the intangible asset and the market environment. It calculates an index representing the state of the intangible asset and places it in a multidimensional latent space. It projects the index onto a human-understandable observation space for visualization. It also observes state transitions along the time axis and generates future states based on a state transition model, presenting them as future scenarios including multiple branches.

[0033] Here, latent space is a representation space inherent in an information processing entity (general-purpose computer, trained model, or information processing system including these), and is a multidimensional space that does not presuppose human perception. This concept is similar to the concept known as "latent space" in the field of machine learning (see Non-Patent Literature 1). Furthermore, observation space is a subspace of state quantities in the latent space that are observed in a form understandable to the user. This concept is similar to the space of observation vectors defined by the measurement equation in a state-space model, and has a structure in which a part of the state space is selectively observed by an observation matrix (selection matrix) that defines a mapping from state vectors to observation vectors (see Non-Patent Literature 2).

[0034] Furthermore, the present invention allows for the external reference, use, or replacement of evaluation modules or calculation modules that can be swapped out according to the purpose of use or evaluation perspective, thereby improving the comparability, transparency, and expandability of evaluation results.

[0035] Here, an externally provided computation model or analysis module (hereinafter referred to as an "external module") refers to a computation function that is provided independently of the information processing system executing this information processing method and is called via a predetermined interface. The external module may be implemented as a service provided over a network (Web service, API, etc.), as a plugin, or as a combination thereof. Input and output to the external module may be performed using industry standard formats such as PMML (Predictive Model Markup Language), XML, JSON, or structured data formats based on a predetermined schema.

[0036] Specific examples of the aforementioned external modules may include prediction models (regression models, classification models, time series prediction models, etc.), optimization algorithms (linear programming, integer programming, constrained optimization, etc.), rule engines (business rules, decision tables, etc.), scoring models (credit scores, risk scores, etc.), or combinations thereof. This information processing method can utilize the external modules by sending input data and receiving output results without knowing the internal structure of the external modules (black box utilization).

[0037] Furthermore, the present invention may include at least the following configurations as configurations that can be added to the basic configuration. Each of these configurations may be implemented in combination with all or part of the basic configuration, or each configuration may be implemented independently. (1) Calibration of evaluations based on the attributes of the evaluator and normative criteria, regression learning of the decision path, and identification and learning intensity control of external observer models of third-party entities based on external observations. (2) Ensuring transparency, including control over the disclosure of decision-making path attributes, third-party audits by auditing bodies, retention of audit records, and detection of tampering through digital signatures, hashes, etc. (3) Integration of publicly available intangible asset information (including not only success stories but also failure stories, and including feature representations contained in trained models) and assignment of metadata to said publicly available information. (4) Multilayer evaluation models, stratified confidence levels, recording and reproducibility of evaluation history, generation of hierarchical explanations of evaluation results, and mutual learning among multiple evaluation systems (distillation learning, associative learning, etc.). (5) A configuration that introduces metrological concepts (measurement resolution, measurement range, error band) into distance evaluation and performs compatibility verification of external modules (scale consistency, resolution consistency, reproducibility, robustness, version consistency, etc.) and acceptance / rejection decisions. In the aforementioned compatibility verification, metadata provided by the external module (scope of application, accuracy information, version information, etc.) is referred to and its consistency with this information processing method is verified. (6) Display of difference comparisons based on the branching point (significance display, development of stratified differences, explanation of difference factors, etc.) and decision support based on difference comparisons. (7) Market entry feasibility assessment, trade partner suitability assessment, calculation of protective investment limits based on expected future profits, and optimization of strategic defense lines based on gate strength correction factors and constrained optimization. [Effects of the Invention]

[0038] According to the present invention, the relationship between intangible assets, protective measures, and the market environment can be treated as a dynamic structure that includes subject dependence, time changes, and uncertainty, and users can explore and compare the transition of states and future scenarios through intuitive operation.

[0039] Furthermore, by incorporating concepts such as measurement resolution, measurement range, or error band into the distance or similarity evaluation results, the robustness, reproducibility, and accountability of judgments in comparisons, rankings, etc., can be improved.

[0040] Furthermore, by making a part of the state transition model or evaluation model replaceable as an external module, flexible operation becomes possible depending on the application, purpose, or evaluation perspective. In particular, by adopting a form in which the external module is called as a service, this information processing method can update, replace, or use multiple modules in combination with the external module without depending on the internal implementation of the external module. [Brief explanation of the drawing]

[0041] [Figure 1] This is a block diagram showing the overall configuration of the information processing system according to the present invention. [Figure 2] This is a block diagram showing the functional configuration of the processing unit. [Figure 3] This is an explanatory diagram showing the data structure of intangible assets. [Figure 4] This is a flowchart showing the overall processing procedure of the information processing method according to the present invention. [Figure 5] This is a conceptual diagram showing the projection relationship between the latent space and the observed space. [Figure 6] This is an example screen showing a display example in the observation space. [Figure 7] This is an example screen showing a future scenario. [Figure 8] This is an explanatory diagram showing examples of measurement resolution and error band assignment. [Figure 9] This is a conceptual diagram illustrating the relationship between calibration and value discrepancy. [Figure 10] This is an example screen showing an example of updating and comparing differences triggered by an external event. [Figure 11] This is an example screen showing an example of visualizing audit records and reliability levels. [Modes for carrying out the invention]

[0042] Embodiments of the present invention will be described below with reference to the drawings. Note that the following embodiments are not limiting to the present invention, and various modifications are possible within the scope of the technical concept of the present invention. [Examples]

[0043] An embodiment of the information processing method according to the present invention will be described below. The information processing method according to this embodiment treats the state in which intangible assets belong to an individual, corporation, organization, or group not as a single concept, but as a state quantity that is established as a result of the combination of multiple asset elements, and calculates an index representing the state of the intangible assets based on the asset elements and their combined state.

[0044] (System Configuration Overview) Figure 1 shows the overall configuration of the information processing system according to this embodiment. The information processing system 1 includes an information processing device 10, a user terminal 20, an external information source 30, and an external arithmetic module 40. The information processing device 10 comprises a processing unit 11, a storage unit 12, and a communication interface unit 13. The user terminal 20 comprises a display unit 21 and an input unit 22.

[0045] Figure 2 shows the functional configuration of the processing unit 11. The processing unit 11 includes a state representation means 100, a visualization means 101, a future state generation means 102, a calibration means 103, and an external observer model 104. The state representation means 100 calculates an index 207 representing the state of intangible assets based on asset elements and their combined states, which will be described later, and performs the process of placing the index in the latent space 210. The visualization means 101 performs the process of converting the calculated index into an observation space 220 that can be understood by humans and displaying it through the display unit 21.

[0046] (Definition of terms) In this specification, "attribution" means that an intangible asset or an asset element constituting an intangible asset is treated as being linked to a specific entity for the purpose of evaluation and index calculation. Such attribution includes both attribution based on legal rights and attribution not based on legal rights. It should be noted that the attribution in this invention is a concept for defining the premise for evaluation and index calculation, and does not determine the attribution of legal rights or responsibilities.

[0047] Furthermore, "subject" refers to the scope set as the target of evaluation or index calculation in this information processing method. This scope may be set by the information processing method based on the user's instructions or requests, or it may be set automatically by prediction based on the user's history and the intangible assets they hold. This scope may include individuals, corporations, organizations, and groups, and may also be set considering regional factors, economic factors, distribution structure factors, etc., depending on the evaluation target. When this subject is used as a criterion for evaluating intangible assets, it is called the "evaluation subject," and when it is treated as the recipient of intangible assets or asset elements, it is called the "recipient subject."

[0048] (Basic definition of intangible assets) In this embodiment, an intangible asset 200 is defined as a state in which multiple asset elements combine to create value. Figure 3 shows the data structure of the intangible asset 200. The intangible asset 200 includes asset elements 201-205, combination status 206, indicator 207, update history 208, and audit log 209. Even if individual asset elements 201-205 belong to the entity alone, those asset elements are not necessarily treated as a state in which the intangible asset 200 has value.

[0049] For example, factors that may influence intangible assets include municipalities, countries or regions, economic alliances or economic zones between countries, and factors related to trade flows and supply chains. In other words, these factors (regional factors, economic factors, distribution structure factors) are treated as external conditions or environmental factors that affect the intangible asset, and their influence is assigned to and combined with the asset element. In this combination, an importance level indicating the ratio of influence may be further assigned and treated in combination.

[0050] As one concrete example, the external conditions or environmental factors act as factors that change the distribution of evaluations in the latent space 210 used for evaluation, and the bias in this distribution is reflected in the trend of the calculated evaluation indicators, and is treated in conjunction with the asset elements. Here, the latent space 210 is a representational space inherent in the information processing entity, as defined in the means for solving the problems of this specification, and is a multidimensional space that does not presuppose human perception.

[0051] This information processing method evaluates information regarding regional factors, economic factors, and distribution structure factors, then maps the evaluation results to corresponding asset elements and combines them with other asset elements to form state quantities that constitute intangible assets 200.

[0052] Thus, the information processing method according to this embodiment calculates an index 207 representing the state of the intangible asset 200 based on the asset element and its combined state 206, and this index 207 is used in subsequent embodiments for observing temporal changes, placing in the latent space 210, and calibrating evaluation criteria.

[0053] (Intangible asset value and market value) In this embodiment, the value of the intangible asset 200 is evaluated as a value that reflects not only a value calculated based on the combined state 206 of the asset elements itself, but also a value adjusted according to the attributes of the valuation entity in the market in which the intangible asset 200 is located (hereinafter referred to as market value). This market value is treated as an adjustment value applied according to the market environment in which the intangible asset 200 is located and the attributes of the valuation entity.

[0054] Here, market value refers to the value estimated based on the assumption that the intangible asset 200 will be used in a specific market or business environment. This market value is treated as potentially varying depending on the attributes of the valuation entity, i.e., the industry, market, business area, etc., to which the entity acquiring or using the intangible asset belongs.

[0055] Furthermore, in Example 4, the concept of market value is separated into general market value 412 and value specific to the valuation body 413, and is described in detail as an explicit operation called calibration.

[0056] (Required binding element) Specifically, in this embodiment, the following asset elements are defined as essential connecting elements for establishing the intangible asset 200 (see Figure 3).

[0057] (Element 1) Asset element 201 that is established as a legal right such as a patent right, utility model right, design right, or trademark right.

[0058] (Element 2) Assets related to know-how such as technical ideas, design information, and business knowledge (Element 202)

[0059] In this embodiment, element 1 (asset element 201) is not the dominant factor as an intangible asset on its own, and is treated as only having value as an intangible asset 200 when combined with element 2 (asset element 202).

[0060] In other words, with respect to patent rights and utility model rights, since the subject matter of these rights is a technical idea, the substance protected by these rights corresponds to know-how, and rights that are not combined with such know-how (asset element 202) are treated to a limited extent as a subject of valuation.

[0061] Furthermore, while design rights differ from patent rights in that the external shape is specified in writing, etc., design rights are treated as having value as intangible assets 200 when the design knowledge related to the external shape is combined as know-how (asset element 202).

[0062] With regard to trademark rights, the value of the intangible asset 200 is formed when the chronological history of market recognition, goodwill, and transaction history is combined as informational elements representing the state of the intangible asset 200. Trademark rights that do not have this history combined may be evaluated as having no value as an intangible asset 200.

[0063] (auxiliary connecting element) Furthermore, in this embodiment, the following elements are defined as asset elements that can be supplementarily combined with the essential combination elements.

[0064] (Element 3) Asset Element 203 relating to trade secret protection capability based on systems, regulations, and actual operational practices for managing and maintaining confidential information.

[0065] With respect to Element 3 (Asset Element 203), the valuation is based not on the existence of formal systems or regulations themselves, but on the ability to continuously operate those systems.

[0066] Furthermore, the following asset elements can be added as auxiliary components.

[0067] (Element 4) Asset element 204 relating to copyrighted works

[0068] With regard to copyrighted works, the value of their existence, whether tangible or intangible, is treated as being calculated based on past performance, and this information processing method does not consider that a sufficient initial value exists for such copyrighted works. This value may be set based on information obtained through user input or integration with external systems.

[0069] (Evaluation modification element: brand value) When the aforementioned element 4 (asset element 204) is combined, its value is greatly influenced by the following valuation modification elements.

[0070] (Element 5) Asset element 205 related to brand value formed from market recognition, credibility, usage history, transaction history, etc.

[0071] The aforementioned element 5 (asset element 205) is evaluated by referring to accumulated information based on past events and is treated as something that is gradually capitalized based on its correspondence with resource inputs such as funds and man-hours.

[0072] On the other hand, asset element 205 related to the brand value is also treated as a state quantity that can be impaired in the opposite direction to its assetization. For example, if another brand is launched in the same or similar market, or if tangible or intangible goods that can be considered similar to the brand begin to circulate in the market, asset element 205 is assessed as being impaired in proportion to the decline in market share. This assessment is carried out in conjunction with the expansion or contraction of the overall market.

[0073] Furthermore, if the brand reaches a stage where it is assessed to have spilled over into markets different from those initially anticipated, the asset component 205 relating to the brand value may be shown as increasing non-linearly.

[0074] (Impact of exercising rights) If asset element 205 relating to the brand value is based on the scope of rights to which the exercise of those rights is left to the discretion of the rights holder, then the actions of the rights holder, whether or not they take legal action or other action to exercise those rights, will be treated as affecting the value of asset element 205.

[0075] This information processing method evaluates the impact of an action on the impairment or increase of intangible asset value based on its similarity to past cases. In this case, the evaluation results at the present time and the evaluation results at a future time can be displayed side by side.

[0076] (Evaluation considering market value) The information processing method according to this embodiment includes a step of calculating an index 207 based on the combination state 206 of the asset elements, and then performing a correction process on the index 207 to reflect the market value. If the index 207 includes element 5 (asset element 205), the correction process shall be applied to the index 207 that reflects the influence of element 5.

[0077] In this adjustment process, an adjustment factor is applied to the intangible asset 200 being evaluated, taking into account the market to which the intangible asset 200 belongs, whether or not the entity utilizing the intangible asset 200 has already entered that market, and the costs, equipment, know-how, etc., required to enter that market.

[0078] For example, if the valuation entity is already in the same or a similar market as the intangible asset 200, the market value may be adjusted to be higher because the additional costs to utilize the intangible asset 200 will be relatively low. On the other hand, if the valuation entity belongs to a different market than the intangible asset 200, the market value may be adjusted to be lower by considering the barriers to entry required to utilize the intangible asset 200.

[0079] Furthermore, the correction process is explicitly defined as calibration in Example 4 and is described in detail as calibration of market value based on the attributes of the evaluation entity (for example, users, organizations, etc., to which attributes such as regional factors, economic factors, and distribution structure factors are assigned).

[0080] (Selection of the evaluator) In the information processing method according to this embodiment, the user can select evaluation criteria corresponding to the evaluating entity. These evaluation criteria can be switched by the user's operation and are not limited to discrete selections; they may also be set continuously using ratios or the like.

[0081] This makes it possible to switch between displaying the value of the same intangible asset (200) from the perspective of its own market and from the perspective of other markets.

[0082] Furthermore, in Example 4, the concept of the evaluation entity is extended to serve as the basis for calculating and calibrating the evaluation entity's intrinsic value 413.

[0083] (Calculation of indicators and placement in latent space) In this specification, "combination state 206" refers to a state in which the correlation, contribution, or dependency between multiple asset elements 201 to 205 is quantified. The combination state 206 is calculated as factor loadings, correlation coefficients, contribution rates, output values ​​of weighted aggregate functions, or combinations thereof, and is expressed as a scalar quantity, a vector quantity, or a matrix.

[0084] In the information processing method according to this embodiment, an index 207 is calculated based on the combination state 206 of the asset elements 201 to 205.

[0085] In quantifying the combined state 206, the correlation, contribution, or dependency between the multiple asset elements 201 to 205 may be calculated as factor loadings, correlation coefficients, contribution rates, or weighted aggregate functions (see Patent Documents 2 and 3). For example, the factor loadings of each asset element for latent factors extracted by factor analysis may be calculated, and an index 207 representing the combined state 206 may be constructed based on these factor loadings. Alternatively, the states of the multiple asset elements 201 to 205 may be calculated as a single scalar quantity or a low-dimensional vector quantity by weighted aggregate operations using the framework of a multi-attribute utility function.

[0086] In the conversion from asset elements 201 to 205 to index 207, the correspondence between asset elements and indexes is not limited to one-to-one. That is, multiple asset elements may contribute to a single index component, and a single asset element may contribute to multiple index components. This many-to-many correspondence is analogous to the relationship between observed variables and factors in factor analysis, or the structure of the measurement model in structural equation modeling.

[0087] The calculated index 207 is placed and held in the latent space 210 of the information processing entity. Figure 5 shows the projection relationship between the latent space 210 and the observation space 220.

[0088] The aforementioned latent space 210 is a representational space inherent in an information processing entity, as defined in the means for solving the problems of this specification, and is a multidimensional space that does not presuppose human perception. This concept is similar to the concept known as "latent space" in the field of machine learning (see Non-Patent Document 1).

[0089] The arrangement of the indices 207 in the latent space 210 may be transformed into a human-understandable observation space 220 based on predetermined rules to facilitate user understanding. The observation space 220 is a subspace of the state quantities in the latent space 210 that are observed in a form understandable to the user, as defined in the means for solving the problems of this specification. This concept is analogous to the space of observation vectors defined by the measurement equations in a state space model, and has a structure in which a part of the state space is selectively observed by an observation matrix (selection matrix) (see Non-Patent Literature 2). In Figure 5, the projection 222 shows the transformation from the latent space 210 to the observation space 220, and the boundary 221 shows a conceptual demarcation within the latent space 210.

[0090] (display) In the information processing method according to this embodiment, the user can select the asset element to be displayed from among the multiple asset elements 201 to 205 that constitute the intangible asset 200. This information processing method displays the index 207 based on the combination state 206 of the asset elements, calculated according to the above description, and the corrected index 207 that reflects the market value, as images on the observation space 220. Figure 6 shows an example of the display in the observation space 220.

[0091] This selection is not limited to being statically set, but may also be configured to be dynamically changed in response to user actions. This allows users to instantly switch between the asset elements to be displayed while checking the status of each asset element.

[0092] Furthermore, users may set multiple asset elements 201 to 205 as a single integrated asset element. In this case, the information processing method calculates an index 207 based on the combined state 206 of the multiple asset elements and generates a display based on an overview perspective. This allows users to visually grasp the value of intangible assets 200 from their own perspective.

[0093] Each calculated asset element or integrated asset element is displayed as an image in the observation space 220. In this embodiment, the image may be displayed as a three-dimensional image so that the user can intuitively grasp the relationships between the asset elements.

[0094] Here, "an image recognizable as three-dimensional" is not limited to images showing only a single elevation or plan view, but rather includes the concept of images that simultaneously show multiple views of an object from different perspectives, specifically, three-view drawings.

[0095] More preferably, the image is displayed as a perspective view and may be displayed with shading or lighting. For example, by using lighting techniques such as loop lighting or Rembrandt lighting, the combination and interrelationships of asset elements can be visually emphasized.

[0096] Furthermore, the image that can be recognized as three-dimensional may be displayed using binocular parallax, thereby allowing the user to grasp the state of the intangible asset 200 as a three-dimensional display with depth.

[0097] Furthermore, the information processing method according to this embodiment may be configured to allow the user to arbitrarily change the position of their viewpoint. This makes it possible for the user to observe the intangible asset 200 from different viewpoints, and to grasp the combined state 206 of the asset elements and its changes from multiple perspectives.

[0098] Furthermore, in the information processing method according to this embodiment, the position of the light source in the lighting may be configured to be arbitrarily changed according to the user's operation. This allows the user to observe the state of the intangible asset 200 under different lighting conditions and to more clearly grasp irregularities, discontinuities, hidden structures, etc., in the combined state 206 of the asset elements.

[0099] Furthermore, the information processing method according to this embodiment may include a step of providing audible notification in response to changes in the display. Such audible notification is output in response to, for example, the occurrence of a predetermined event during automatic playback along the time axis 320, or changes in state in response to time specification operations or seek operations by the user. This allows the user to grasp changes in the state of the intangible asset 200 through auditory information in addition to changes in the visual display.

[0100] Furthermore, the information processing method according to this embodiment includes a step for receiving instructions regarding time. The user can specify whether to display information corresponding to a past, present, or future point in time. In response to the instructions regarding time, the information processing method updates the indicators 207 of each asset element or combination state 206 and changes the displayed content. This allows the user to confirm the changes in the state of the intangible asset 200 over time, based on the perspective they desire.

[0101] (Expansion of experiential expression) Furthermore, the information processing method according to this embodiment may include a step of presenting the state of the index 207 in the observation space 220 to the user through senses other than sight.

[0102] Presentation through senses other than sight may include one or more of the following:

[0103] (Auditory presentation) In addition to the aforementioned audible notifications, the state changes of indicator 207, the probability of branching scenarios occurring, or the risk level may be auditorily represented as pitch, volume, timbre, rhythm, or harmonic structure. For example, by representing an increase in the evaluation indicator as an ascending melody and a decrease as a descending melody, users can recognize state changes without relying on visual cues.

[0104] (Tactile presentation) The state of the indicator 207 in the observation space 220 may be presented to the user via a haptic feedback device. This haptic feedback may be expressed as a vibration pattern, temperature change, pressure change, or texture sensation. For example, by expressing an increase in the value of the intangible asset 200 as a warming sensation and a decrease in value as a colding sensation, or by expressing the uncertainty of the branching scenario as the intensity of vibration, the user can grasp the state experientially.

[0105] (kinesthetic presentation) The state of the indicator 207 in the observation space 220 may be presented to the user via a force feedback device. This force feedback may be expressed as resistance, repulsion, or traction. For example, by expressing the bonding strength between intangible assets 200 as resistance during operation and the influence from the market environment as an external force, the user can intuitively grasp the relationships between the states.

[0106] (Olfactory presentation) The state of the indicator 207 in the observation space 220 may be presented to the user via a fragrance output device. This olfactory presentation may be expressed as a fragrance pattern associated with a specific state or classification. For example, associating asset elements that are showing a growth trend with a specific fragrance can help maintain attention during long-term observation.

[0107] (Multisensory presentation) Each of the above sensory modalities may be used individually or in combination. By combining multiple sensory modalities, complex states that are difficult to perceive with a single sense can be presented to the user more intuitively. Furthermore, users with impairments in one or more senses can be provided with equivalent information through other senses.

[0108] Thus, the information processing method according to this embodiment may include a step of expressing the index 207 in the observation space 220 experientially through one or more of the five human senses. This experiential expression may be used in place of visual visualization, or in combination with visual visualization.

[0109] Furthermore, the specific device configurations for realizing presentation through each of the aforementioned sensory modalities (such as tactile feedback devices, force feedback devices, and fragrance output devices) can be those of known devices, and the technical scope of the present invention is not limited to such device configurations.

[0110] (System Configuration) As shown in Figure 1, the information processing system 1 that implements the information processing method according to this embodiment comprises at least a processing unit 11 that performs arithmetic processing, a storage unit 12 that stores information, a display unit 21 that displays information to the user, and an input unit 22 that receives input from the user.

[0111] The processing unit 11 executes each of the steps described in this embodiment based on the information stored in the storage unit 12 and the input from the user.

[0112] The storage unit 12 stores information about the multiple asset elements 201 to 205 that constitute the intangible asset 200, information about the combination state 206 of the asset elements, information about the valuation entity, information about the market value, and information about the calculated index 207. The storage unit 12 also holds information about the arrangement of the index 207 in the latent space 210.

[0113] The display unit 21 displays information representing the combination state 206 of asset elements in the observation space 220, generated by the processing unit 11, in a format that can be visually grasped by the user. The display unit 21 may be configured to display an image that can be recognized as three-dimensional.

[0114] The display unit 21 may include, in addition to or instead of, a visual display, an auditory output means (speaker, earphones, etc.), a tactile output means (vibrator, temperature change element, etc.), a force-feedback means (force-feedback device, etc.), or an olfactory output means (fragrance generator, etc.). This allows the information processing system 1 to present information through the most appropriate sensory modality according to the user's sensory characteristics or usage environment.

[0115] The input unit 22 accepts user input such as the selection of asset elements to be displayed, the selection of the evaluation body, instructions regarding time, changes in the viewpoint, and other operational instructions.

[0116] As shown in Figure 2, the processing unit 11 includes at least a state representation means 100 and a visualization means 101.

[0117] The state representation means 100 calculates an index 207 representing the state of the intangible asset 200 based on the asset elements 201 to 205 and their combined state 206, and executes a process to place the index 207 in the latent space 210. The state representation means 100 treats multiple asset elements, including asset element 201 (element 1) relating to legal rights, asset element 202 (element 2) relating to know-how, asset element 203 (element 3) relating to trade secret protection ability, asset element 204 (element 4) relating to copyright, and asset element 205 (element 5) relating to brand value, as representations in a multidimensional space.

[0118] The visualization means 101 converts the index 207 calculated by the state representation means 100 into a human-understandable observation space 220 and performs the process of displaying it through the display unit 21. The visualization means 101 performs operations such as switching the asset elements to be displayed, switching the evaluation entity, and changing the viewpoint in response to operations by the user through the input unit 22.

[0119] The observation space 220 is not limited to a display format, but can be configured as a representational format different from the latent space 210, to assist human understanding and judgment.

[0120] The processing unit 11 may, if necessary, include a future state generation means 102 that generates a future state using the current state represented by the state representation means 100 as an initial condition. Details of the future state generation means 102 will be described in subsequent embodiments.

[0121] Furthermore, the processing unit 11 may include extension means that, if necessary, allow at least a portion of the calculation processing to be replaced or used in combination with an externally provided calculation model or analysis module (hereinafter referred to as the external module 40).

[0122] The external module 40 is provided independently of the information processing system 1 and is a arithmetic processing function that is called via a predetermined interface (see Figure 1). Any of the following configurations or combinations thereof may be used for cooperation with the external module 40.

[0123] (a) Service calls over the network If the external module 40 is provided as a Web service or API (Application Programming Interface), the information processing system 1 sends a request to the external module 40 via the network through the communication interface unit 13 and receives the calculation result as a response.

[0124] (b) Plugin-in If the external module 40 is provided as a plug-in, the information processing system 1 dynamically loads the plug-in and performs calculations via a predetermined interface.

[0125] (c) Collaboration in batch processing format When the external module 40 is provided as batch processing, the information processing system 1 provides the input data to the external module 40 as a file or data stream, and obtains the output result after processing is complete.

[0126] The input / output data format for the external module 40 may be PMML (Predictive Model Markup Language), XML, JSON, CSV, or a structured data format based on a predetermined schema. By using these standard data formats, general-purpose collaboration that is independent of the provider of the external module 40 can be achieved.

[0127] Specific examples of the external module 40 include the following:

[0128] (i) Predictive models Machine learning models or statistical models such as regression analysis models, classification models, time series forecasting models, neural networks, and decision trees.

[0129] (ii) Optimization algorithms Optimization techniques such as linear programming, integer programming, constrained optimization, and genetic algorithms.

[0130] (iii) Rule engine Rule-based inference capabilities such as business rules, decision tables, and decision graphs.

[0131] (iv) Scoring Model A model for calculating evaluation scores such as credit score, risk score, and suitability score.

[0132] (v) Simulation engine Probabilistic simulation functions such as Monte Carlo simulation and agent-based simulation.

[0133] This information processing system 1 can utilize the external module 40 in a manner that involves transmitting input data and receiving output results (black box utilization) without knowing the internal structure of the external module 40. This allows the provider of the external module 40 to provide computational functions to this information processing system 1 while keeping the internal implementation of the module confidential.

[0134] Furthermore, the aforementioned extension means does not outsource the calculation of the evaluation index itself, but is used in a configuration that assists the processing by the processing unit 11.

[0135] This information processing system 1 may be configured as a single device, or as a system in which multiple devices are connected in a communicative manner. In the latter case, the components of the processing unit 11, storage unit 12, display unit 21, and input unit 22 may be distributed across different devices.

[0136] The functions of the processing unit 11 in this information processing system 1 may be realized by a program that causes a computer to execute each of the steps described in this embodiment. This program may be recorded on a computer-readable recording medium or distributed via a communication line.

[0137] (Connection to subsequent embodiments) The concepts of "latent space 210" and "observation space 220" defined in this embodiment will be used in subsequent embodiments as a basis for observing state transitions along the time axis (Embodiment 2), generating and branching future states (Embodiment 3), and calibrating evaluation criteria (Embodiment 4).

[0138] Furthermore, the concept of "asset elements 201-205" in this embodiment will be expanded in subsequent embodiments as a relationship with "asset properties" in a multidimensional space. Here, "asset properties" are computational degrees of freedom set to describe asset elements in latent space 210, and are concepts equivalent to factors in factor analysis or latent variables in structural equation modeling. The asset elements 201-205 and the asset properties do not need to correspond one-to-one; a single asset element may be represented by multiple asset properties (e.g., multiple columns of a factor loading matrix), and multiple asset elements may be represented by a single asset property (e.g., a scalar quantity calculated by weighted aggregation) (see Patent Documents 2 and 3). Through this many-to-many correspondence, the combination state 206 of asset elements can be computably represented as a position or distribution in latent space 210.

[0139] (effect) The information processing method according to this embodiment acquires information on the multiple asset elements 201 to 205 and their combined state 206 as described above, calculates an index 207 for each asset element and its combined state, and displays the index 207 on the same observation space 220. This allows the user to visually grasp the combined state 206 of the asset elements constituting the intangible asset 200 and its changes over time. [Examples]

[0140] (Examples of display and operation involving time-series transitions) The following describes another embodiment of the information processing method according to the present invention.

[0141] This embodiment is based on the information processing method described in Embodiment 1, and calculates an index 207 based on predetermined calculation criteria from information concerning an intangible asset 200 and the rights protecting said intangible asset 200, and makes said index 207 displayable and operable over time.

[0142] In this embodiment, while maintaining the concept of the latent space 210 introduced in Embodiment 1, a configuration is shown that allows for the observation of the temporal change of the index 207 in that space.

[0143] (Information acquisition) The information processing method according to this embodiment includes a step of acquiring information on multiple intangible assets 200 and corresponding rights held by an individual or legal entity (company or organization). The acquired information includes the duration of the rights, the registration date, the scope of the rights, the usage status, and information on the usage environment and market conditions of the intangible assets 200.

[0144] Furthermore, in this embodiment, if external environmental information such as treaties between nations, national budgets, customs duties, tax systems, administrative guidelines or administrative operations, or reports published by government agencies may affect the status of the intangible asset 200 or rights, such external environmental information may be acquired as information used in calculating the indicator. This may improve the accuracy of the representation regarding valuation in international transactions, etc.

[0145] Furthermore, it is also possible to acquire IR information from companies and investors in proportion to the national budget, and use their investment plans as external environmental information for calculating indicators.

[0146] Furthermore, when it is deemed important to consider the impact of municipal-level external environmental information as information used in calculating one of the elements constituting Intangible Assets 200 (for example, element 2) (i.e., when it is of higher importance than others), the indicator may be calculated assuming that there are periods when the impact of these factors outweighs that of other external environmental information. This calculation method can improve the accuracy of representations for small-scale retailers, service businesses, restaurants, etc.

[0147] This information may be obtained through user input, or through automated acquisition using RPA (Robotic Process Automation), etc. Furthermore, information regarding current financial transactions and market conditions, various rights and obligations, publicly available information from other companies, international affairs, etc., may be obtained as discrete automated information acquisition at set time intervals by using APIs, etc. (see Figure 1).

[0148] This allows, for example, when it is foreseen that damage to intangible asset 200 will exceed a predetermined threshold due to an event unexpected by the user (such as an accident, disaster, or dispute), the system to automatically acquire information and push a detailed report to the user (see Figure 10). In other words, the information processing method may be configured to acquire additional information when conditions specified by the user are met.

[0149] It should be noted that the aforementioned external environmental information indicates the external conditions of the subject of evaluation and is different in nature from the publicly available intangible asset information (reference knowledge used for evaluation) introduced in Example 6.

[0150] (Calculation of indicators based on calculation criteria) Next, based on the acquired information, an index 207 is calculated that indicates the status of each intangible asset 200 and each right. In this embodiment, the index 207 is calculated based on predetermined calculation criteria and is used to enable comparison between different intangible assets 200 or their status at different points in time.

[0151] The calculation criteria may be set based on physical quantities or definitive numerical values ​​if the information can be treated as such. For example, if a legal right has a term of duration, an evaluation function indicating the status of the right can be constructed using the publication date, the expiration date, or the time from the publication date to the expiration date.

[0152] On the other hand, if the calculation includes factors that are difficult to treat as definitive physical quantities, such as whether or not pension payments will be made in the future, objections, revocation proceedings, or the stability of rights based on the relationship with prior art, or market trends, the calculation criteria may be set based on statistical methods, probabilistic methods, or a predetermined reference model.

[0153] In this embodiment, as an example of such calculation criteria, an information processing model that has learned information about the legal system or business practices under which the right is established can be used as a reference criterion. In this case, the information processing model may reflect case law and trade practices in a particular country or region, and the weighting of the referenced information may be adjusted according to the area of ​​operation of the company or organization and its relevance to the applicable law.

[0154] The aforementioned reference model may be used to assess the correspondence between the current state of intangible assets 200 or rights and similar past cases. This correspondence will be dealt with in more detail as a similarity rate, which will be detailed in subsequent embodiments.

[0155] However, in this embodiment, the reference space for arranging the index 207 is treated as fixed. That is, even if the index 207 is calculated based on different points in time, different intangible assets 200, or different evaluation conditions, it can be compared with each other by being placed on the same reference space.

[0156] Therefore, in practice, it is preferable to use the same model as the information processing model used as the reference criterion. Examples of information processing models include LLMs (Large Language Models). This makes it possible to suppress the discontinuous changes in the relative relationships between the indices 207 due to fluctuations in the criteria.

[0157] While the above explanation primarily focuses on legal rights, similar calculation criteria and indicators will be applied to information regarding the scope of rights, usage, the usage environment of Intangible Assets 200, and market conditions.

[0158] (External observation instrument function and historical record documents) In this embodiment, the information constituting the reference model may include the following:

[0159] (a) Standards and guidelines Documents referenced as normative standards (e.g., UN Charter, laws and regulations, industry guidelines). These are detailed in Example 4 as components of standard module 410.

[0160] (b) Historical documents A collection of documents containing records of past successes and failures related to 200 intangible assets and rights. This information is used to constitute the external monitoring function.

[0161] The aforementioned historical documents may be used as information that constitutes the external observation device function.

[0162] The external observation function may also function as a reference basis for positioning the current status of the intangible asset 200 and rights relative to other intangible assets 200, related market information, historical events, rights status, or combinations thereof present in the latent space 210.

[0163] For example, when using a pre-trained language model, the set of data relating to past intangible assets 200, rights, economic events, or market events used to train the language model may correspond to the set of historical documents that constitute the external observation function.

[0164] In other words, the training data in a trained language model functions as a collection of past cases for evaluating the current state of intangible assets 200 or rights.

[0165] In addition, the external observation function is embodied in Example 4 as an external observation model 104 that identifies internal responses from input and output for a specific third-party entity (see Figure 2).

[0166] (Placement in latent space and transformation into observation space) The calculated index 207 is placed in the latent space 210 described in Example 1 (see Figure 5).

[0167] In this embodiment, a process is performed to convert the arrangement of the indicators 207 in the latent space 210 into the observation space 220 that can be understood by the user. This conversion process generates a display on the observation space 220.

[0168] (Data structure of update history) In the information processing method according to this embodiment, the update history 208 of the index 207 is stored in a format having the following data structure (see Figure 3).

[0169] (a) Update identifier An identifier for uniquely identifying each update operation. This identifier may consist of a sequential number, a UUID (Universally Unique Identifier), or an identifier based on a timestamp.

[0170] (b) Update time The date and time the update was performed. The update time is recorded as Coordinated Universal Time (UTC) or based on the time zone specified by the user.

[0171] (c) Index value before update The value of index 207 before the update. This value can be expressed as a scalar quantity, a vector quantity, or a matrix, and is recorded as a coordinate in latent space 210.

[0172] (d) Updated index value The updated value of index 207. It is recorded in the same format as the index value before the update.

[0173] (e) Update factors Information indicating the factors that triggered the update. These factors may include any of the following: - Time elapsed (regularly updated) - User input - Information obtained from external source 30 - Receipt of external trigger information 500 (see Figure 10) - Correction by calibration means 103 (see Figure 2)

[0174] (f) Update difference The difference between the index value before the update and the index value after the update is 501 (see Figure 10). This difference is calculated and recorded as distance, direction vector, or rate of change in the latent space 210.

[0175] (g) Reference update source Reference information to external sources 30, external modules 40, or other intangible assets 200 related to the update.

[0176] The update history 208 is stored in the storage unit 12 defined in Embodiment 1 and made accessible upon user request. Furthermore, the update history 208 works in conjunction with the audit log 209 to ensure traceability of updates (see Figure 11).

[0177] (Generating the display) In this embodiment, multiple intangible assets 200 and indicators 207 showing the status of their corresponding rights are placed on a display space configured as an observation space 220 (see Figure 6). This allows users to visually grasp the relative relationships between the intangible assets 200 and rights at a given point in time.

[0178] The generation of the aforementioned display only needs to express the state or relative relationship of the indicator 207 in a manner that is easy for the user to understand. For example, the state of the elements constituting the intangible asset 200 may be associated with the state of the land, and the boundary 221 surrounding the land may be associated with the state of rights or access routes.

[0179] Boundaries can be represented by, for example, castle walls, fences, moats, stone walls, ditches, mountain ranges, cliffs, rivers, oceans, etc. The number or difficulty of access routes to the intangible asset 200 may also be represented by density, width, or shade.

[0180] Furthermore, the paths through which the value of the intangible asset 200 may be impaired may be dynamically represented, for example, by relating them to insect damage. In this case, the priority of countermeasures can be indicated by the number of insects, the amount of movement, the intensity of the colors, etc.

[0181] (Display updates along the timeline) Furthermore, in this embodiment, the display is updated in accordance with the status of the intangible assets 200 and rights, which may change over time. The user can select or switch the point in time to be displayed via the operation input unit 22 corresponding to the time axis 320 (see Figure 6).

[0182] The updates may be continuous updates along the time axis 320, or discrete updates corresponding to predetermined time points.

[0183] In response to this operation, the position, spacing, or spatial relationship of each indicator 207 placed on the observation space 220 is updated, making it possible to compare the state at different points in time on the same observation space 220. This allows users to understand, continuously or intermittently, how the state of intangible assets 200 and rights changes over time.

[0184] In updating the display along the time axis 320, the index values ​​at each point in time recorded in the update history 208 are referenced, and the state at that point in time is reproduced. This allows the user to check the state of the intangible asset 200 at any point in the past on the observation space 220.

[0185] (Highlighting of differences) Furthermore, in this embodiment, the user can operate the system to highlight the differences between the state at a specific point in time and the state at a past or future point in time (see Figure 10). This makes it easier to visually recognize the direction and degree of change over time.

[0186] Furthermore, to convey transitions to the past or future, visual techniques such as fade-out, fade-in, zoom-out, and zoom-in may be used, or combined with other visual techniques, to enhance understanding and awareness.

[0187] In highlighting the differences, the update differences recorded in the update history 208 are referenced. These update differences are visually displayed as differences 501 on the observation space 220, allowing users to intuitively understand which asset elements have changed and to what extent.

[0188] (Connection to subsequent embodiments) In this embodiment, the index 207 is displayed in an observable form along the time axis 320, but behind this display, the latent space 210 introduced in Embodiment 1 exists.

[0189] In subsequent embodiments, the latent space 210 is explicitly treated as a space that is difficult for humans to directly understand, and the control of projection 222 from that space to the observation space 220, as well as the generation of multiple future states and the presentation of branching scenarios (future scenarios 300), are described in detail (see Figure 7).

[0190] Furthermore, the concepts mentioned in this embodiment as correspondences with past examples using a reference model are developed in subsequent embodiments as state transition models based on similarity rates.

[0191] (effect) According to this embodiment, an information processing method is provided that allows observation of state changes along the time axis 320 while maintaining the index calculation and display configuration defined in Embodiment 1.

[0192] This allows users to understand the changes in the status of intangible assets 200 and the rights protecting such intangible assets 200, both continuously and discretely, from the past to the present and into the future.

[0193] Furthermore, highlighting differences based on a specific point in time makes it easier to visually recognize the direction and extent of changes over time.

[0194] Furthermore, by placing indicators 207 at different points in time on a fixed reference space, comparability between time points is ensured, making it possible to objectively evaluate the value fluctuations of intangible assets 200.

[0195] Furthermore, by maintaining the update history 208 in a structured data format, the traceability of changes in indicator 207 is ensured, making it possible to refer to the status and basis for changes at each point in time during audits or verifications.

[0196] Note that the display mode, operation method, and specific configuration of the indicator 207 in this embodiment are examples only, and various modifications are possible within the scope of the information processing method described in Embodiment 1. [Examples]

[0197] (Examples of future state generation, branching scenario presentation, and projection control from latent space to observed space) The following describes another embodiment of the information processing method according to the present invention.

[0198] This embodiment is based on the information processing method described in Embodiment 1, which calculates an index 207 from information relating to intangible assets 200 and the rights associated therewith, and further extends the observation and manipulation of state transitions along the time axis 320 described in Embodiment 2 to include the generation of future states, the presentation of multiple branching scenarios, and the control of projection 222 from the latent space 210, which is difficult for humans to directly understand, to the observation space 220, which is understandable to humans (see Figure 5).

[0199] In this embodiment, the concept of the latent space 210, which was introduced in Example 1 and unfolded on the time axis 320 in Example 2, is explicitly treated as a multidimensional latent space 210, and the relationship between the latent space 210 and the observation space 220 is explained in detail.

[0200] (Definition of terms) In this embodiment, the "asset elements" constituting the intangible asset 200 or rights refer to the constituent units (elements 1 (201) to 5 (205)) defined in Example 1, and have the following two properties.

[0201] Firstly, the aforementioned asset elements include parts that are obvious through registration, etc. (the real part of the legal asset elements). Examples include the description of claims in patent rights and the indication of registered trademarks in trademark rights.

[0202] Secondly, the aforementioned asset elements include non-obvious components (conceptual constituent units, equivalent to imaginary parts). For example, goodwill, know-how, brand value, or even legal asset elements may be accompanied by non-obvious elements such as the actual market value of the right, synergies with other rights, and ease of exercise.

[0203] The aforementioned non-trivial portion is described in the latent space 210, and the user perceives it in the observation space 220, which is transformed from the latent space 210 by a predetermined projection rule 222 or dimensionality reduction rule (see Figure 5).

[0204] Furthermore, since legal asset elements also have non-obvious elements, all asset elements include the elements described in the latent space 210.

[0205] In this embodiment, "asset property" represents a computational degree of freedom set for describing the asset element in the latent space 210. There is no one-to-one correspondence between the asset element and the asset property; multiple asset elements may be represented by a single asset property, and a single asset element may be represented by multiple asset properties.

[0206] (1) Calculation of indicators and placement in latent space The information processing method according to this embodiment includes a step of acquiring information concerning intangible assets 200 and rights associated therewith held by an individual or legal entity (company or organization). The acquired information may include asset elements (201 to 205) constituting the intangible asset 200, the state of connection between elements 206, the duration of the rights, the registration date, the scope of the rights, the usage status, and external environmental information surrounding the intangible asset 200 and rights (see Figure 3).

[0207] Next, based on the acquired information and the historical record documents (standard documents and historical documents) defined in Example 2, an index 207 representing the status of each intangible asset 200 and each right is calculated.

[0208] (1-1) Integrated processing of asset elements In calculating the aforementioned index 207, the obvious parts of the asset elements (parts that can be confirmed by registration, etc.) and the non-obvious parts (value components that are difficult to observe, sensitivity to the external environment, reputation, potential risks, etc.) are integrated and processed as asset properties in the potential space 210.

[0209] The aforementioned process may include, for example, one of the following methods:

[0210] (a) A method that represents the trivial part as a real number component and the non-trivial part as an imaginary number component, and integrates them using complex number operations.

[0211] (b) A method that uses the membership function of a complex fuzzy set to represent uncertainty in terms of amplitude and phase.

[0212] (c) A method for representing unobservable interference effects using complex probability amplitudes in quantum-like modeling.

[0213] (d) Methods for representing non-trivial value components as the intensity dimension (axis perpendicular to the plane of the paper) in the intellectual capital model.

[0214] (e) A method that represents verifiable components and components requiring estimation as different subspaces in a multidimensional vector space and integrates them using tensor operations.

[0215] The selection of the aforementioned method may be determined according to the nature of the intangible asset 200 to be evaluated, the characteristics of the available data, or the user's requirements.

[0216] (1-2) Placement into latent space The calculated indicators 207 are arranged and held in the multidimensional latent space 210, which is not intended to be visually perceived by the user (see Figure 5). The latent space 210 is not limited to a three-dimensional space or time axis 320 that humans can directly understand, and the relationships between multiple indicators 207 may be expressed as positional relationships or distance relationships within the latent space 210.

[0217] (2) External observation instrument function and relativization processing In this embodiment, the external observation function introduced in Embodiment 2 is used to relatively position the current status of intangible assets 200 and rights.

[0218] (3) Generation of future states and branching events based on similarity ratio Furthermore, in this embodiment, future states are generated using a state transition model based on the similarity rate between the history of the indicator 207, the current state of the intangible asset 200 and its associated rights, and past success or failure events related to the intangible asset 200 and rights included in the historical document group constituting the external observer function (see Figure 7).

[0219] The similarity rate may be calculated based on the correspondence between an index 207 representing the current state of the intangible asset 200 or rights and an index 207 representing past events related to the intangible asset 200 or rights.

[0220] (3-1) Method for calculating the similarity ratio The similarity rate is calculated using statistical methods. Specifically, the similarity rate may be calculated by checking the correlation coefficient for the observation space 220 presented by this information processing method or presented or modified by the user. The observation space 220 is defined by the input group, the output group, and the time axis 320.

[0221] In calculating the similarity rate, the setting of the population (the range of past examples to be compared) and the evaluation method (correlation coefficient, cosine similarity, Euclidean distance, etc.) affect the calculation result. This information processing method provides pre-set default values ​​for the population and evaluation method. These default values ​​may include a setting in which the group of examples included in the publicly available intangible asset information introduced in Example 6 is used as the population and Pearson's correlation coefficient is used as the evaluation method. Users may use the default values, or they may change the population or evaluation method based on their own judgment. If a user makes a change, this information processing method records the changed settings and retains them in a reproducible format.

[0222] (3-2) Conditions for generating branching events The state transition model may be configured to generate a future state branching event (state transition branch 301) when the similarity rate exceeds a predetermined threshold (see Figure 7). The threshold may be set as a probability value, and a calculation method similar to the next token selection probability in a language model may be used.

[0223] For example, if the future state with the highest probability of occurrence is within a predetermined high-probability range (e.g., 80-90%), only a single future state may be generated. If the probability of occurrence of that state decreases and falls within a predetermined probability range (e.g., in the 70% range), a second future state may be generated as a branching scenario.

[0224] In other words, the threshold acts as a boundary of the probability of occurrence that controls the display of branching scenarios, and multiple branches become visible when the probability of the maximum likelihood scenario falls below the threshold.

[0225] (3-3) Future state generation using external modules At least a portion of the state transition model or future state generation in this information processing method may be performed using an externally provided computation model or analysis module (hereinafter referred to as external module 40) (see Figure 1).

[0226] The generation of future states using the external module 40 may include the following steps:

[0227] (a) Preparation of input data This information processing method converts the current placement status of indicators 207, past update history 208, external environment information, and prediction conditions into a predetermined data format (JSON, XML, PMML, etc.) and prepares them as input data for the external module 40.

[0228] (b) Calling external modules This information processing method transmits the prepared input data to an external module 40 via an API call over the network, a plug-in interface, or a batch processing interface.

[0229] (c) Receiving the output results This information processing method receives output results from an external module 40 that include an indicator 207 representing a future state, branch probability, confidence score 601, or a combination thereof.

[0230] (d) Integration of output results This information processing method arranges the received output results as future states in the latent space 210 and subjects them to the projection 222 and display processing described in (4) and subsequent sections.

[0231] Specific examples of the external module 40 may include the following:

[0232] (i) Time series forecasting service A service that takes historical time-series data of 207 indicators as input and predicts future values ​​for 207 indicators. For example, it provides time-series forecasting models such as ARIMA, Prophet, and LSTM.

[0233] (ii) Scenario generation service A service that generates multiple future scenarios (300) with probabilities, using the current state and external conditions as input. For example, a service that provides Monte Carlo simulations.

[0234] (iii) Risk assessment services A service that takes the status of an intangible asset 200 as input and outputs a risk score or risk scenario related to that intangible asset 200.

[0235] (iv) Market forecasting services A service that uses market environment data and the status of 200 intangible assets as input to predict future market value or market trends.

[0236] This information processing method can utilize the external module 40 based solely on the relationship between input and output, without knowing its internal structure (algorithm, training data, parameters, etc.). This configuration allows the provider of the external module 40 to provide a predictive or evaluation function to this information processing method without disclosing its internal implementation, which would be a source of competitive advantage.

[0237] Furthermore, this information processing method may call multiple external modules 40 in parallel and compare or integrate the output results obtained from each external module 40. The comparison or integration may be performed based on the framework of collaborative evaluation between multiple systems, which is detailed in Example 8.

[0238] The uncertainty (confidence interval, variance, confidence score 601, etc.) included in the output of the external module 40 may be presented to the user as an error band 400 or a confidence score 601 representation in the display in the observation space 220 described in (4) and later (see Figures 8 and 11).

[0239] (4) Projection of multidimensional branching structures and construction in observation space The aforementioned future states and branching scenarios may be generated in the latent space 210 as a multidimensional state transition structure. To facilitate user understanding, this multidimensional state transition structure may be projected 222 onto the human-readable two- or three-dimensional observation space 220 based on a predetermined projection 222 rule or dimensionality reduction rule (see Figure 5).

[0240] The projection rule 222 is a rule for transforming the arrangement of the indicators 207 in the multidimensional latent space 210 into a two- or three-dimensional space that is perceptible to humans. The projection 222 transforms the complex relationships in the latent space 210 into a form that can be visually grasped.

[0241] Specific examples of the projection 222 rule or dimensionality reduction rule may include principal component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), Uniform Manifold Approximation and Projection (UMAP), multidimensional scaling (MDS), or the encoder portion of an autoencoder. The selection of these methods may be determined according to the number of dimensions of the latent space 210, the nature of the relationships between the indices 207 (linearity, nonlinearity, preservation of local structure, etc.), or the user's requirements. Furthermore, multiple projection 222 rules may be configured to be switchable, allowing the user to select the display format.

[0242] For example, the structure after projection 222 may be configured as a root-like structure of a tree that branches out toward the future starting from the current state, and may be represented as a tree diagram, a fishbone, or a similar structure (see Figure 7).

[0243] Here, the structure is generated as a result of projecting a multi-dimensional state transition structure in the latent space 210 onto the two-dimensional or three-dimensional observation space 220, and should not be simply understood as a planar tree diagram.

[0244] In addition, each branch path may be expressed as the thickness, width, brightness, transparency, depth representation, or angular dominance rate of a line according to the occurrence probability of the future state (see Fig. 7).

[0245] (5) Current location display by view control and floating map In this embodiment, the user may determine their viewing position using the floating map window 310 (see Fig. 6). The floating map window 310 map - represents the current state and the branch structure in the observation space 220, and may be capable of magnification, reduction, rotation, bird's - eye view, and viewpoint movement.

[0246] This operation may provide an operability similar to the viewpoint operation in, for example, Google Map or Google Earth.

[0247] Thereby, the user can observe the future state while grasping which position of the multi - dimensional branch structure the current state corresponds to.

[0248] In the floating map window 310, a cursor or blinking display indicating the current position of the user may be arranged in front of the branch point or at the intersection. Also, the plurality of transition paths branching from the branch point may be displayed as lines having a thickness or width according to the occurrence probability, and the future state with a higher occurrence probability has a more visible display mode.

[0249] By doing so, the user can receive support from this information processing method regarding which future state should be preferentially observed.

[0250] Furthermore, by giving depth to the root-like structure and using expressions that mimic atmospheric perspective and depth of field, the user's current position in a three-dimensional or multidimensional space can be displayed more effectively, improving understanding.

[0251] In the aforementioned root-like structure, the two axes other than the time axis 320 may be arranged with asset elements (201-205) constituting the intangible asset 200, sets of elements, types of rights, or other indicators 207, which may be predetermined or may be changed based on user specifications.

[0252] The information processing method according to this embodiment may be configured to select the future scenario 300 with the highest probability of occurrence as the first display based on asset elements predetermined or selected by the user, and to display the first display so that it becomes the dominant field of view in the user's field of view. Other future states are displayed with a field of view or display ratio corresponding to their respective probability of occurrence, and these display ratios may be dynamically changed if specified by the user.

[0253] (6) Providing reasons in response to a request for explanation The information processing method according to this embodiment is configured to display the reason for the branching scenario presented in response to a request from the user. The reason does not need to be displayed at all times in the initial state, and may be displayed only when the user shows interest in the branching scenario.

[0254] The reasoning statement may be generated by the information processing method, referencing the historical document set constituting the external observation function, and based on the similarity relationship with the current intangible asset 200 and rights status. Alternatively, the reasoning statement may be generated via an external information processing device or an external API (external module 40).

[0255] In other words, the system may use past success or failure events as a motif to show which past case the current state is similar to, and explain that a branching event (state transition branch 301) occurred because the similarity (similarity rate) exceeded a predetermined threshold.

[0256] For example, when a user performs actions such as mouseover, long press, or right-click, or when they perform actions such as gaze fixation, hand gestures, or pointing in an AR / VR environment, an explanatory window may be displayed, showing the reason why the branching scenario was generated.

[0257] This structure allows users to gain a deeper understanding of why a particular branching scenario was presented. Furthermore, this explanatory function is useful not only for evaluating intangible assets 200 and intellectual property, but also for educational purposes related to economics and management.

[0258] (7) AR / VR environment and time axis manipulation The information processing method according to this embodiment is not limited to a 2D screen environment, but may also be performed in an AR or VR environment. In an AR / VR environment, the observation space 220, branching scenario (future scenario 300), floating map window 310, and explanatory window are arranged in three-dimensional space and may be operated by gaze, gestures, proximity operations, etc.

[0259] Furthermore, in addition to the configuration described in Example 2, the operation of the time axis 320 may be performed using a spatial rail, a curved timeline, a jog dial, or a spatial representation of the time axis 320 (see Figure 6).

[0260] Furthermore, for visual representations during transitions to the past or future (fade out, fade in, zoom out, zoom in, etc.), the same representations as those described in Example 2 may be used.

[0261] (8) Clarification of the relationships between the examples This embodiment integrates and expands the concepts of Embodiment 1 and Embodiment 2 as follows:

[0262] Relationship with Embodiment 1: The concept of "latent space 210" introduced in Embodiment 1 is elaborated as a multi-dimensional space in this embodiment. Also, the concept of "asset elements" (201-205) in Embodiment 1 is developed as the relationship with "asset properties" in this embodiment.

[0263] Relationship with Embodiment 2: The concept that was treated as a state change along the time axis 320 in Embodiment 2 is generalized as a multi-dimensional state transition structure in the latent space 210 in this embodiment. Also, based on the concepts of the external observer function, reference specification documents, and history document groups introduced in Embodiment 2, this embodiment formulates the generation of future states by the "similarity rate".

[0264] Uniqueness of this embodiment: In this embodiment, by introducing an explicit conversion process of projection 222 from the latent space 210 to the observation space 220, a technical means is provided to convert multi-dimensional information that is difficult for humans to directly understand into a visually graspable form (see Figure 5).

[0265] (9) Effects As described above, according to this embodiment, an information processing method is provided that holds information regarding the intangible asset 200 and the rights associated therewith multi-dimensionally in the latent space 210, and projects 222 the latent space 210 to the observation space 220 that is understandable by humans.

[0266] Thereby, the following effects are obtained:

[0267] (a) Ease of understanding multi-dimensional information By projecting 222 the multi-dimensional latent space 210, which is difficult for humans to directly understand, to the two-dimensional or three-dimensional observation space 220, complex relationships can be visually grasped (see Figures 5 and 6).

[0268] (b) Prioritizing branching scenarios The relative importance of multiple future states can be intuitively understood through the field of view dominance rate, line thickness, etc., corresponding to the probability of occurrence (see Figure 7).

[0269] (c) Decision support through explainability By explaining the reasons for the presented branching scenarios in terms of similarities to past cases, transparency in decision-making is ensured.

[0270] (d) Deepening understanding through perspective recognition The floating map window 310 allows users to always be aware of their location within the multidimensional branching structure, thereby supporting their understanding of the overall picture (see Figure 6).

[0271] (e) Expandability to AR / VR environments By supporting display environments that are not limited to two-dimensional screens, it becomes possible to provide a more immersive observation experience.

[0272] (f) Integration of Example 1 and Example 2 Building upon the basic concepts of Example 1 and the time-series development of Example 2, this approach adds new technological value in the form of future prediction and branching prediction in multidimensional space.

[0273] Users can recognize their own perspective, understand multiple branching scenarios for the future, and grasp the reasons why those branches were presented (see Figure 7). [Examples]

[0274] (Examples of calibration of evaluations based on evaluator attributes and normative criteria, regression learning of decision paths, and identification enhancement of external observer models)

[0275] The following describes yet another embodiment of the information processing method according to the present invention.

[0276] (0) Positioning of this embodiment and relationships between embodiments This embodiment is constructed based on the following embodiment.

[0277] Direct premise (required): In Example 1, the concept of market value (separated into general market value 412 and subject-specific value 413 in this example), the definition of asset elements (201-205), and the introduction of the latent space 210 are implemented (see Figures 3 and 5).

[0278] In Example 2, the concept of a standard document (extended as standard module 410 in this example), the concept of an external observer function (embodied as external observer model 104 in this example), and a time-series transition observation function are introduced.

[0279] In Example 3, the explicit definition of the latent space 210, the concept of similarity rate, the state transition model, and the projection 222 onto the observation space 220 are described in detail (see Figure 5).

[0280] Indirect premise (foundation): The asset elements (elements 1 (201) to 5 (205)) in Example 1 form the basis for evaluating the synergy effect of the valuation subject's inherent value 413 in this example (see Figure 3).

[0281] Uniqueness of this embodiment: This embodiment includes the introduction of an explicit concept called calibration, the extension of the evaluation subject attributes to include decision path attributes, predictability evaluation by regression learning of decision paths, the concretization of the external observer function into the external observer model 104, the improvement of the accuracy of the external observer model 104 through sequential exploration and collection, focusing on specific targets by selecting the load distribution of learning intensity, and the introduction of an evaluation method based on the substantive effects of international law and treaties, etc. (see Figures 2 and 9).

[0282] Effects on subsequent embodiments: The concepts introduced in this embodiment are further developed and applied in the following embodiments. Embodiment 5 involves the disclosure control of decision path attributes (targeting the decision path attributes introduced in this embodiment), Embodiment 6 involves the dynamic learning of the external observer model 104 (based on the sequential search and collection introduced in this embodiment), Embodiment 7 involves the application of the external observer model 104 to inter-subject influence evaluation, Embodiment 8 involves the confidence level 601 evaluation of the evaluation model (based on the calibration concept of this embodiment), Embodiment 9 involves the retrospective application of the measurement resolution 401 to the distance evaluation of this embodiment, and Embodiment 10 involves the use of evaluation subject attributes to adjust the protection investment coefficient (see Figures 8 and 11).

[0283] This embodiment is based on the calculation method for intangible assets 200 and indicators 207 related to rights described in Embodiment 1, the observation of time-series transitions described in Embodiment 2, and the projection 222 control from the latent space 210 to the observation space 220 described in Embodiment 3, and is an example in which the calibration of the evaluation criteria in the information processing method is explicitly performed by the calibration means 103 (see Figures 2 and 9).

[0284] This embodiment demonstrates that the index 207 calculated and displayed in Examples 1 to 3 can be calibrated along two axes: (i) its value to whom (dependent on the evaluator) and (ii) what criteria are used (dependent on norms). Furthermore, it includes a configuration that (iii) performs regression learning using the user's decision-making process as an attribute, and (iv) generates an external observer model 104 (a concrete manifestation of the external observer function conceptually introduced in Example 2) that identifies internal responses to third-party entities other than the user based on external observations, and allows its learning intensity to be enhanced by the user's selection (load distribution) (see Figure 2).

[0285] (1) Intangible assets as the object of measurement and the need for calibration Unlike physical quantities (length, mass, time, etc.), the value, creditworthiness, or risk of intangible assets 200 and rights are measures for which there is no absolute and unchanging standard, or the standard itself is subject to change depending on the environment and the assessor.

[0286] In the case of physical measuring instruments, calibration against known standards is essential to ensure the accuracy of measurement results. Similarly, in the valuation of intangible assets 200 and rights, comparing valuation results at different points in time or between different entities without explicitly calibrating the valuation standards will result in measurements that lack validity.

[0287] The information processing method according to this embodiment is based on the arrangement of the index 207 in the latent space 210 introduced in Embodiment 1, and when projecting the index 207 onto the observation space 220 222, the calibration means 103 performs calibration of the evaluation criteria from the following two viewpoints (see Figures 5 and 9).

[0288] (A) Calibration of market value based on the attributes of the evaluator The value of the item being evaluated is calibrated to determine "for whom" it holds value and what that value represents. This calibration allows for the calculation of the value discrepancy (Δ) 411 between the general market value 412 and the value specific to the valuation body 413 (see Figure 9).

[0289] (B) Calibration of credit and risk assessments based on normative standards The system calibrates the level of reliability or risk associated with the evaluation target, based on "what criteria." This calibration is performed as a distance evaluation from the reference module 410 (see Figure 9).

[0290] (2) Calibration of market value based on the attributes of the evaluator

[0291] (2-1) General market value and value specific to the valuation subject The information processing method according to this embodiment treats the concept referred to as market value in Embodiment 1 by more explicitly separating it into general market value 412 and value specific to the valuation subject 413 (see Figure 9).

[0292] The aforementioned general market value 412 is an evaluation value assuming an unspecified number of market participants, and is a value obtained by calculating the basic indicator 207 in Example 1.

[0293] On the other hand, the aforementioned value specific to the valuation entity 413 is a valuation value for a specific valuation entity (individual or legal entity (company or organization)) that uses the information processing method, and is a value that fluctuates depending on attributes such as the intangible assets already possessed by the valuation entity, business area, geographical conditions, and strategic policies.

[0294] This separation formalizes the dependence of market value on the valuation subject, which was implicitly suggested in Example 1, as an explicit, computable concept.

[0295] (2-2) Acquisition of evaluation subject attribute information (static attributes and judgment path attributes) In this embodiment, the information processing method includes a step of acquiring attribute information of the evaluation subject. The attribute information may include at least the following static attributes.

[0296] In other words, this may include the composition of intangible assets 200 and rights already held by the valuation entity, the valuation entity's main business areas and markets, the valuation entity's geographical scope of activity, the valuation entity's strategic policy or investment policy, and the valuation entity's organizational capabilities or resources.

[0297] Furthermore, in this embodiment, the attribute information may also include the decision path formed by the user when using the information processing method (the sequence of operations such as the order, selection, reference, comparison, and weighting used in this information processing method).

[0298] In other words, path information—which information a user referenced, in what order, and by what criteria, and which candidates they adopted or rejected—can be defined as attribute information that represents the decision-making characteristics of the evaluator.

[0299] The aforementioned decision-making process may be captured, for example, as log or serial data, including: the viewing order, time spent, focus, history of switching calibration conditions (evaluator attributes or normative criteria) for the displayed evaluation groups, selection of comparison targets, similarity search conditions, filter conditions (time range, industry, region, etc.), weighting operations for indicator 207 (value, credit, risk, etc.), alternative generation, history of counterfactual simulation execution, and final decisions (adopt, hold, reject) and justification inputs.

[0300] The aforementioned decision path may be encoded as a sequence feature in the latent space 210 and represented as a decision path attribute (see Figure 5). In this way, the evaluation subject attribute is treated not only as static profile information but also as a recursive element that can change over time.

[0301] Furthermore, users may be allowed to set weighting parameters for each element constituting the judgment path attributes. This allows users to emphasize or reduce specific judgment elements (e.g., emphasis on value assessment, emphasis on risk assessment, time urgency, etc.) according to their own strategy or tactics.

[0302] (2-3) Regression learning of decision-making pathways and evaluation of predictability (long-term calibration) The information processing method according to this embodiment may evaluate the predictability of the user by using the consistency between the judgment output corresponding to the judgment path (forecast, investment judgment, partnership judgment, rights exercise policy, etc.) and the results observed after the passage of time (actual results, success or failure, profit or loss, dispute occurrence, change in credit, etc.).

[0303] Specifically, the information processing method may take the decision path at decision time t0 as input, and use the result label obtained at time t1 after a predetermined period (e.g., several weeks to several months) as a training signal or reinforcement signal to update the weights of the decision path attributes.

[0304] This makes it possible to adjust the evaluator's inherent value 413 (correction of the evaluator's evaluation criteria) and calculate the confidence level of the judgment 601 (predictability score) based on whether the user's own judgment is consistent with reality (see Figure 11).

[0305] Furthermore, since results including failures (inconsistencies) can be incorporated as learning targets to correct the user's judgment tendencies, this information processing method values ​​not only success cases but also failure cases, contributing to the improvement of long-term decision-making ability.

[0306] This configuration is based on the time-series transition observation function in Example 2, and explicitly handles the temporal relationship between the decision point and the result observation point.

[0307] Furthermore, the learning of the decision-making process is, in principle, performed using anonymized, aggregated, or concealed features, and information that could identify individual users is removed.

[0308] (2-4) Distillation know-how and metacognition through learning decision-making pathways from multiple users If the information processing method according to this embodiment is capable of accumulating and learning the decision paths of a large number of users, the information processing method may go beyond the decision paths unique to each individual user to extract highly predictable decision path patterns and model them as distilled know-how.

[0309] For example, the decision paths of a group of users with high predictability scores are clustered to form high predictability decision types (hereinafter also referred to as know-how clusters). The distance between a user's current decision path and the know-how cluster is then calculated in the latent space 210 in Example 3, and the proximity is displayed in the observation space 220 (see Figures 5 and 6).

[0310] (Integration with Example 9: Consideration of measurement resolution) Furthermore, the calculation of the aforementioned distance is subject to the constraint of the measurement resolution 401, which will be detailed in the subsequent Example 9 (see Figure 8). That is, if the difference in the calculated distance is less than the measurement resolution 401, multiple know-how clusters or user decision paths may be treated as identical with "no significant difference." This suppresses excessive subdivision of decisions based on apparent minor differences.

[0311] When the distance is small, the user's judgment can be interpreted as being consistent with the distillation know-how, while when the distance is large, it can be interpreted as having a tendency to deviate.

[0312] In this case, the information processing method may, as metacognition, provide, for example, the following comparative displays: consistency (distance) with the user's decision-making process over a specific past period (e.g., the past 12 months), consistency (distance) with the majority decision-making process of the population (whole or specific segment) at the same point in time, and consistency (distance) and estimated predictability (probability) with high predictability clusters.

[0313] Here, the population may be specified not only by time range, but also segmented and compared based on conditions such as industry, region, company size, and business phase.

[0314] Furthermore, for past samples that made similar judgments, it is also possible to present the time-series changes (judgment trajectory and outcome progression) showing how those judgments were updated over time and what results they led to.

[0315] The presentation may be presented as a result based on statistical processing (frequency, distribution, regression, etc.), or as a probabilistic estimation (probability of success, probability of risk occurrence, expected loss, etc.) using an information processing model (e.g., LLM (Large Language Models)) as mentioned in Example 2.

[0316] (2-5) Analysis of synergy effects and unique risks (calculation of value discrepancy) The information processing method according to this embodiment analyzes the synergy effects or unique risks when the intangible asset 200 or right to be evaluated is incorporated into the existing asset group of the evaluation entity, based on the arrangement relationship in the potential space 210 (see Figures 5 and 9).

[0317] Specifically, the following processes may be performed in the latent space 210.

[0318] (a) Calculation of general market value vector The evaluation target is placed in a reference space that assumes an unspecified number of market participants, and a vector V_market representing market value is calculated. This vector corresponds to the general market value of 412 (see Figure 9).

[0319] (b) Calculation of the value vector intrinsic to the evaluator The objects to be evaluated are placed in a space that takes into account the existing asset group and judgment path attributes of the evaluation entity, and a vector V_entity representing the value unique to the evaluation entity is calculated. This vector corresponds to the evaluation entity's unique value 413 (see Figure 9).

[0320] (c) Calculation of value discrepancy The discrepancy Δ = |V_market - V_entity| between the two vectors is calculated. This discrepancy is displayed in the observation space 220 as the value discrepancy (Δ) 411 (see Figure 9).

[0321] In calculating the aforementioned discrepancy, it is necessary to consider the measurement resolution 401 detailed in Example 9 (see Figure 8). That is, if the calculated discrepancy is less than the measurement resolution 401, it is treated as "no significant difference" between the general market value 412 and the value specific to the valuation body 413.

[0322] If the aforementioned deviation Δ exceeds a predetermined threshold and is a significant difference exceeding the measurement resolution 401, the information processing method may calibrate the evaluation result for the evaluation entity using the calibration means 103 and present a corrected evaluation value that differs from the general market value 412 (see Figures 2 and 9).

[0323] (2-6) Specific application examples For example, consider a case where an appraisal entity that operates a retail chain evaluates the location value (a type of intangible asset 200) associated with commercial real estate in a certain area.

[0324] From the perspective of a large, unspecified number of market participants, properties located in existing commercial districts are generally considered to have a higher market value.

[0325] However, if the evaluating entity's strategy involves developing new market areas and it already has its own logistics network and existing stores nearby, then a different property located in a generally less-valued area may actually be of higher value to that entity.

[0326] This is because elements 2 (202: know-how) and 3 (203: trade secret protection ability), as defined in Example 1, generate value that is not valued in the general market when combined with the existing assets of the valuation entity (see Figure 3).

[0327] Furthermore, if the evaluation body has a history of successful or unsuccessful decision-making processes in similar site evaluations, the attributes of that decision-making process can be regressively reflected in the calibration of the current evaluation. That is, if a decision-making process similar to past discrepancies (failures) is detected, the information processing method may display a warning and present additional comparisons (results of similar samples).

[0328] (2-7) Identification of external observer models for third-party entities (regression outside the user) The information processing method according to this embodiment, in addition to regression learning based on the user's own decision-making process, can also set a third-party entity other than the user (such as a country, company, organization, or individual) as the target of observation, and identify the entity's internal response (input / output characteristics) to a limited extent from the information expressed externally to that entity.

[0329] This configuration implements the "external observer function," which was conceptually introduced in Example 2, into a concrete computable model as the external observer model 104 (see Figure 2).

[0330] (Collection and preprocessing of input / output information) Specifically, for the entity under observation S, an externally observable output sequence O(t) is collected from official announcements, publicly available documents, behavioral history, transaction history, public events such as lawsuits and public notices, and market reactions.

[0331] Furthermore, the input sequence I(t) may be estimated as an input proxy (external conditions such as regulatory environment, capital policy, supply and demand, partnerships, and investments) based on publicly available information.

[0332] A step may be added to assign weights to the information constituting the input proxy in advance. These weights may be arbitrary (e.g., uniformly the same value) or pre-trained values ​​as initial values, and the accuracy of the external observer model 104 can be improved by regressively updating them based on subsequent results.

[0333] Furthermore, confidential information to which access permissions have been granted may be used as needed.

[0334] (Identification and construction of response models) The information processing method may involve performing an inverse transformation (inverse estimation) from the input and output based on the degree of correlation or causal consistency of the collected multiple events (I(t), O(t)), and identifying a response model M_S that represents the internal processing of the subject S being observed.

[0335] The response model M_S may be held as a parametric representation on the latent space 210, or it may be configured as an External Observer Model 104 using the information processing model (e.g., LLM) mentioned in Example 2 (see Figure 2). Other methods such as Bayesian estimation and discrimination models may also be applied.

[0336] The external observer model 104 specifically implements the concept of "external observer function" introduced in Example 2, and provides a function to dynamically construct training data corresponding to past record documents (history documents) from publicly available information from third parties.

[0337] (Functionality of the response model) The response model M_S may include the following functions:

[0338] (a) Predictive function The predicted future output for the subject S is calculated. This predicted value may be accompanied by uncertainty (confidence interval, variance, confidence 601 score, etc.) (see Figure 11).

[0339] (b) Explanatory function It includes an explanatory model that outputs the feature contributions that form the basis of the prediction, allowing it to present the reasoning behind the decision to the user.

[0340] (c) Warning function If the uncertainty exceeds a predetermined threshold, the information processing method may display a recommendation to collect additional data or a warning to the user to postpone a decision.

[0341] (d) Visualization integration function The external observation instrument model 104 may be used to provide users with a function that allows for highlighting of displays (images visible to the user in the observation space 220) according to the contribution of specific information (see Figure 6). Examples of this highlighting include color, brightness, line width, blinking period, and animation speed. This function makes it easier for users to grasp the causal relationship with the information, thereby enhancing their understanding.

[0342] (3) Calibration of credit and risk assessments based on normative standards

[0343] (3-1) Introduction of standards and norms information Similar to valuation, it is necessary to clearly calibrate the measurement baseline when evaluating aspects such as creditworthiness, governance, compliance, and the appropriateness of exercising rights.

[0344] In this embodiment, standard reference information (hereinafter referred to as standard module 410) that serves as the benchmark for evaluation is introduced (see Figure 9).

[0345] The aforementioned standard module 410 is an extension of the concept of "standard document" mentioned in Example 2, and may be configured as a standard vector or standard area generated by analyzing the United Nations Charter, national laws and regulations, industry guidelines, corporate ethics codes, or codes of conduct adopted by the evaluation body itself.

[0346] (3-2) Evaluation methods based on substantive effects in international law and treaties, etc. A key technical feature of this embodiment is that, with respect to international law, the UN Charter, statements, and treaties exchanged between countries (hereinafter collectively referred to as "international law, etc."), the evaluation method emphasizes the substantive effect of how agreements between parties and related parties were handled, rather than determining whether requirements were met through formal legal interpretation methods (literal interpretation, logical interpretation, etc.).

[0347] (Technical challenges in handling international law, etc.) Regarding international law and other similar laws, interpretations that they take precedence over domestic law and interpretations that they are subordinate to domestic law may coexist, and the degree of legal binding force is often not unambiguously determined. For this reason, methods that determine whether "the requirements are met or not" or "whether or not an effect occurs" using formal legal interpretation and then calculate distance based on the results may not be appropriate in assessing substantive credibility or risk.

[0348] In this embodiment, the following evaluation criteria are introduced to address this technical challenge.

[0349] (Evaluation criteria based on actual effects) The information processing method according to this embodiment will be evaluated with respect to international law, etc., based on the following evaluation criteria.

[0350] (a) The process and basis for the formation of the promise This study analyzes the process by which the relevant international laws were established (negotiation process, agreement-making process, etc.) and the preconditions that formed the basis of those agreements (historical background, economic situation, security considerations, etc.).

[0351] (b) Legitimacy and preconditions We will assess the extent to which the promise is perceived as legitimate between the parties, and whether the preconditions for guaranteeing the promise (mutual trust, balance of interests, existence of a third-party oversight mechanism, etc.) are maintained.

[0352] (c) Discrepancy in interpretation between the parties The extent to which there is or was a discrepancy in interpretation of the wording used in the agreement exists between the parties is analyzed. The discrepancy in interpretation may be quantified as a distance on the latent space 210 (see Figure 5).

[0353] (d) Changes in interpretation and their temporal process This study analyzes the chronological changes in how the parties have interpreted the agreement, as well as the motivations behind those changes (political intentions, economic interests, security needs, etc.). These changes are represented by the time-series transition observation function described in Example 2.

[0354] (e) The period during which the promise was fulfilled (the period during which the actual effect was achieved) This measures the period during which the promise was effectively fulfilled, that is, the period during which the parties actually acted in accordance with the promise. It uses the period during which the actual effect was observed, rather than the formal effective period, as the basis.

[0355] (Structure of the practical effectiveness evaluation model) This information processing method constructs a model for evaluating the substantive effects of international law, etc., based on the evaluation axes (a) to (e) described above (hereinafter referred to as the substantive effect evaluation model).

[0356] The aforementioned effective performance evaluation model performs the following process.

[0357] (1) Calculation of the degree of deviation from the wording The degree of discrepancy between each wording (article, statement, etc.) used in international law and the actual actions of the parties is calculated. This degree of discrepancy may be expressed as a distance in the latent space 210 (see Figure 5).

[0358] Specifically, the obligations described in the articles are set as a reference vector, the actual actions of the parties are represented as action vectors, and the distance between the two is calculated. This distance is processed by the calibration means 103 as a deviation from the reference module 410 (see Figures 2 and 9).

[0359] (2) Evaluation of time-series stability The degree of deviation from the aforementioned wording is evaluated as changing over time. If the degree of deviation remains stable over time (stays within a certain range), the promise is considered to be substantially kept. If the degree of deviation tends to increase over time, the credibility of the promise is considered to have decreased.

[0360] The aforementioned time-series stability may be visualized by the time-series transition observation function in Example 2.

[0361] (3) Construction of a future prediction model A model is constructed to predict future deviations based on the degree of deviation from past wording and its time-series stability. This model may be implemented as the state transition model in Example 3 (see Figure 7).

[0362] This future prediction model estimates the future distance (degree of deviation from the wording) based on the assumption that "the promise will be similarly observed (or not observed) in the future."

[0363] (Addressing the ambiguity of the concept of trust) In this embodiment, we explicitly recognize that the term "trust" is ambiguous and take the following technical measures.

[0364] Instead of treating "credibility" as a single indicator 207, it will be expressed as multiple components based on the aforementioned evaluation axes (a) to (e). The substantive effect (degree of deviation from the wording, time-series stability) will be the primary indicator 207, and the presence or absence of formal legal effect will be the secondary indicator 207. Users will be able to select which evaluation axes to emphasize, and this will be clearly stated as a calibration condition for the evaluation results.

[0365] (Specific application example) For example, consider the case of evaluating a bilateral investment treaty (BIT).

[0366] Formally, the agreement is assumed to have been ratified in both countries and to be in force under domestic law. However, this scenario assumes that one of the parties has amended its domestic law after the conclusion of the agreement in a manner contrary to the agreement, thereby effectively restricting the rights of foreign investors.

[0367] Under traditional formal legal interpretation methods, the agreement may be deemed valid, and investors may be considered to be protected under the agreement.

[0368] However, the actual effectiveness evaluation model in this embodiment performs the following evaluations: As a deviation from the wording, the deviation between the investment protection guaranteed by the agreement's wording and the actual application of domestic law is calculated as a distance. As time-series stability, the trend of the deviation over time after the agreement's conclusion is observed and an expanding trend is detected. As a future forecast, assuming that the expanding trend will continue, further deviations in the future are predicted.

[0369] Based on these evaluation results, this information processing method will inform users that the substantive reliability of the protection under the said agreement has decreased and will warn them of this as an investment risk.

[0370] (3-3) Distance evaluation between action vector and reference module The information processing method according to this embodiment arranges the behavioral data or rights exercise status of the entity to be evaluated (intangible asset 200 or rights holder) as behavioral vectors in the latent space 210 (see Figure 5).

[0371] Next, the distance or directional difference between the action vector and the reference module 410 is calculated (see Figure 9).

[0372] When international law or the like is involved, the reference module 410 is configured as a reference vector based on the wording of the international law or the like, and the distance is calculated as the degree of deviation from the wording detailed in (3-2) above.

[0373] In calculating the aforementioned distance, it is necessary to consider the measurement resolution 401 detailed in Example 9 (see Figure 8). That is, if the calculated distance is less than the measurement resolution 401, it is treated as "no significant discrepancy" between the action vector and the reference module 410.

[0374] If the aforementioned distance expands over time and represents a significant change exceeding the measurement resolution of 401, or if the subject of evaluation exhibits different behaviors depending on the attributes of the trading partner (e.g., strong / weak party, transaction size), the information processing method may calculate this as credit risk or norm deviation risk.

[0375] This risk assessment, when combined with the time-series transition observation function in Example 2, can be expressed as consistency or trend in behavior from the past to the present.

[0376] (3-4) Detection of differences in behavior based on opponent attributes A particularly important function in this embodiment is the detection of cases where the object being evaluated exhibits different behaviors depending on the attributes of the other party.

[0377] For example, if a rights holder does not exercise their rights against large corporations but actively exercises them against small and medium-sized enterprises or individuals, this difference in behavior can be expressed as a dispersion or multimodality of the behavior vectors in the latent space 210 (see Figure 5).

[0378] This information processing method detects such behavioral heterogeneity and visualizes potential credit risk by evaluating it as a deviation from standard module 410 (for example, a norm that emphasizes the fairness of transactions).

[0379] Similarly, when a country exhibits different compliance attitudes towards international law, etc., depending on the attributes of the other country (economic power, military power, political influence, etc.), the aforementioned substantive effectiveness evaluation model calculates the degree of deviation for each attribute of the other country and visualizes the heterogeneity.

[0380] (4) Dynamic switching of calibration conditions, sequential exploration and acquisition, and control of learning intensity (load distribution)

[0381] (4-1) Switching the attributes of the evaluator The information processing method according to this embodiment may allow the evaluation subject attributes to be dynamically changed as parameters.

[0382] This redefines the "selection of evaluation body" function introduced in Example 1 as an explicit concept of calibration.

[0383] Users can perform the following operations: displaying their own company's (current evaluation entity) evaluation value, displaying another company's (hypothetical evaluation entity) evaluation value, and displaying the difference between the two (value deviation (Δ) 411) (see Figure 9).

[0384] Furthermore, in this embodiment, since the evaluation subject attribute includes the decision path attribute, a comparison of cases where the decision path differs even if the evaluation subject is the same (e.g., the difference between an exploratory path and a conservative path) may be displayed.

[0385] (4-2) Switching of normative standards Similarly, the information processing method according to this embodiment may allow the applicable reference module 410 to be dynamically changed (see Figure 9).

[0386] For example, the following options are available: evaluation based on the UN Charter, evaluation based on the laws of a specific country, evaluation based on industry guidelines, evaluation based on the company's own corporate ethics code, and evaluation based on a substantive effectiveness evaluation model (based on the degree of deviation from the wording of international law, etc., and time-series stability).

[0387] This allows users to immediately see how the credit risk or norm deviation risk being assessed changes when the applicable norms change.

[0388] In particular, by comparing and displaying evaluations based on formal legal interpretation with evaluations based on substantive effect assessment models, especially regarding international law, it is possible to visualize the discrepancy between form and substance.

[0389] (4-3) Specifying the population (time, industry, region, etc.) and displaying the distance The information processing method according to this embodiment may allow the user to specify which population to be compared.

[0390] The population may be defined by, for example, the following conditions: a time range (e.g., the past 6 months, the past 3 years, before and after a specific event), industry classification (e.g., manufacturing, IT, healthcare, etc.), regional classification (e.g., by country, by prefecture, by economic zone), and attributes such as company size, stage of growth, and regulatory intensity.

[0391] The information processing method may calculate the distance between the user's current decision path and the majority decision path (or representative trajectory) of the population and display it on the observation space 220 (see Figure 6). This allows the user to understand their distance from the "general market decision at that point in time."

[0392] (4-4) Presentation of time-series changes of similar samples and probabilistic output The information processing method according to this embodiment may extract past samples (decision path, input information at the time, and subsequent results) that are similar to the user's decision path, and present the time-series changes of said samples.

[0393] For example, the period after the decision could be divided into predetermined windows (1 month, 3 months, 12 months, etc.), and the trajectory of how the sample updated its perception of the situation and what results it arrived at could be presented.

[0394] Furthermore, the presentation may be presented as statistically aggregated results (success rate, average profit / loss, variance, risk occurrence rate, etc.), or as probabilities estimated by language models, etc. (probability of future event occurrence, conditional success probability, etc.).

[0395] This function applies the time-series transition observation function from Example 2 to a set of similar cases.

[0396] (4-5) Sequential search and collection (enhanced by external observation) and learning intensity parameters The information processing method according to this embodiment may include a process of sequentially searching for, collecting, and updating input and output events related to the subject S being watched in order to improve the accuracy of the external observer model 104 described in (2-7) (see Figure 2).

[0397] Sequential exploration and collection may include, for example, expanding the scope of public information sources (medium, language, geography, time period), increasing the frequency of event extraction (shortening the collection cycle), adding external variables used to estimate the input proxy, verifying events (consistency checks, counter-searches, deduplication), and updating the model based on the collection results (parameter updates, retraining, distillation).

[0398] In this embodiment, the information processing method may allow setting parameters to control the learning intensity (such as search depth, data collection frequency, verification intensity, and update frequency). Setting a higher learning intensity may consume more computing, search, and data collection resources.

[0399] This configuration enables the dynamic expansion and updating of the "past record document set" that was statically defined in Example 2.

[0400] (Controlling model updates) In the aforementioned model update, the update may be performed if the degree to which a new event contradicts the existing model exceeds a threshold. However, if the confidence level of the event is below the threshold (601), the update may be suppressed and additional data collection may be prioritized (see Figure 11).

[0401] Alternatively, the process may prioritize the search for information sources or events that maximize information gain. Here, information gain is calculated as the amount of entropy reduction or uncertainty reduction in the predictive distribution.

[0402] (4-6) Load allocation as an option for learning intensity (focusing on specific subjects) The information processing method according to this embodiment may provide the user with the option to specify a particular subject S to be observed and adjust the load distribution to improve the learning intensity of the external observer model 104 related to that subject, or to add resources that can handle a new load (see Figure 2).

[0403] Here, load balancing refers to a change in at least one of the following: for example, search depth, validation intensity, model update frequency, model capacity, or allocated computing resources. Resources can be implemented as, for example, compensation, credits, budget limits, processing priority, or additional contractual options.

[0404] Even if the resources of the information processing method provider (for-profit or public enterprise) are limited, users can choose to focus on specific targets, thereby improving the performance of the external observer model 104 with respect to those targets.

[0405] This can be a means of meeting the specific needs of users when an entity that is a low priority for general market participants may hold high strategic importance to a particular user (e.g., a potential trading partner, competitor, or potential partner).

[0406] Furthermore, the additional resources obtained through this load allocation can be reallocated to improve the generalization performance of the 104 external observation model groups (improving the overall prediction function). This makes it possible to achieve both the resolution of individual user issues and the improvement of the overall performance of the information processing method.

[0407] (4-7) Display of Calibration Results The calibration results may be displayed on the observation space 220 as a projection 222 onto the observation space 220 (see Figures 5, 6, and 9).

[0408] Specifically, the following display methods are possible: Display the general market value 412 and the value specific to the valuation body 413 in different colors or brightness levels. Visualize the distance from the reference module 410 using dynamic representations such as "insect damage" in Example 2. Present the change in evaluation values ​​before and after calibration using the "highlighting" function mentioned in Example 2.

[0409] Furthermore, the distance between the decision-making process and the distillation know-how cluster is represented by metaphors such as "fortress walls" and "boundaries" in Example 2. The distance from the population majority is mapped and displayed to visual attributes such as color, brightness, line width, blinking period, and animation speed. The causal relationships of the information are visualized through highlighting (color, brightness, line width, blinking period, etc.) according to the contribution of the external observer model 104.

[0410] Furthermore, the time-series progression of the degree of deviation from the wording of international law, etc., is visualized using the 320-scale time axis representation in Example 2 (see Figure 6). Future predictions based on the substantive effect evaluation model are presented as branching scenarios (future scenario 300) in Example 3 (see Figure 7).

[0411] (5) Clarification of the relationships between the examples This embodiment is constructed based on the concepts of Embodiments 1 to 3, as described below.

[0412] Relationship with Example 1: In this embodiment, the concept of "market value" introduced in Example 1 is separated into "general market value 412" and "valuation subject-specific value 413," and the relationship between the two is explicitly expressed as a calibration operation (see Figure 9). Furthermore, "selection of valuation subject" in Example 1 is redefined in this embodiment as switching of valuation subject attributes and is further expanded to include judgment path attributes. The asset elements (elements 1 (201) to 5 (205)) defined in Example 1 are used in this embodiment as a basis for explaining the synergistic effect of valuation subject-specific value 413 (see Figure 3).

[0413] Relationship with Example 2: The concept of a "standard document" introduced in Example 2 is extended in this embodiment as a "standard module 410," forming the basis for calibration based on standard criteria (see Figure 9). Furthermore, the "external observer function" conceptually mentioned in Example 2 is materialized in this embodiment as an external observer model 104 constructed from input and output observations by a third party (see Figure 2). The time-series transition observation function in Example 2 is used in this embodiment for time-delay alignment of decision paths and results (predictability assessment) and for time-series presentation of similar samples.

[0414] Relationship with Example 3: The latent space 210 detailed in Example 3 is used in this embodiment to arrange value vectors, action vectors, and decision path attributes, and functions as a field for distance evaluation and cluster (distilled know-how) formation (see Figure 5). Furthermore, the projection 222 onto the observation space 220 in Example 3 is used in this embodiment as a means of visualizing calibration results and distance displays. The concept of "similarity rate" in Example 3 is applied in this embodiment to the evaluation of the similarity of decision paths and the matching of input and output patterns of the external observer model 104.

[0415] Furthermore, the distance or similarity measure used to calculate the similarity rate can be selected by the user from cosine similarity, Euclidean distance, Mahalanobis distance, Manhattan distance, correlation coefficient, kernel function-based similarity, etc., depending on the purpose (e.g., improving accuracy, improving computational efficiency).

[0416] Uniqueness of this embodiment: In this embodiment, by introducing the explicit concept of "calibration," a means of technically handling the dependency of evaluation criteria (by whom and based on what criteria) is provided. Furthermore, it is unique in that it incorporates the user's decision-making process as an attribute and regressively calibrates it based on consistency with the results, and introduces an external observer model 104 that identifies internal responses from input / output observations by a third party, making it updatable through sequential exploration and collection, and allowing the learning intensity to be controlled by load distribution selection (see Figure 2).

[0417] In particular, this embodiment introduces an evaluation method for international law and treaties, etc., based on substantive effects (the process of agreement formation, legitimacy, deviations in interpretation, performance record, etc.) rather than formal legal interpretation, and uses the degree of deviation from the wording and chronological stability as the main indicators 207, thereby offering a novel technical concept not found in conventional technology.

[0418] This approach does not output a single "correct evaluation value," but rather presents a set of evaluation values ​​that change depending on the selection of evaluation criteria, thereby improving predictability over time for both the user and the third-party entities that are their trading partners.

[0419] Development into subsequent embodiments: The concepts introduced in this embodiment are further developed and applied in the following embodiments. Embodiment 5 deals with the disclosure control and auditing functions of decision path attributes. Embodiment 6 deals with the integration of public information as a dynamic learning foundation for the external observer model 104. Embodiment 7 deals with the application of the external observer model 104 to inter-entity influence assessment. Embodiment 8 deals with the multi-layered evaluation structure as the theoretical foundation for the calibration concept. Embodiment 9 deals with the explicit specification of the measurement resolution 401 in distance evaluation (retroactively applied to the distance evaluation in this embodiment) (see Figure 8). Embodiment 10 deals with the application of evaluation entity attributes to strategic defense posture optimization.

[0420] (6) Effects As described above, the following effects can be obtained according to this embodiment.

[0421] (a) Explicitly clarifying the dependence on the evaluator The value of the object being evaluated, and for whom, can be quantitatively shown as the discrepancy between its general market value 412 and the value specific to the valuation body 413 (see Figure 9).

[0422] (b) Explicit expression of norm dependence The degree of reliability of the evaluation target, "based on what criteria," can be quantitatively shown as its distance from the standard module 410 (see Figure 9).

[0423] (c) Visualization of hidden value and risks It can detect synergistic effects or unique risks specific to the evaluator that are often overlooked in general assessments, thereby supporting decision-making.

[0424] (d) Detection of behavioral heterogeneity It can detect entities that exhibit different behaviors depending on the attributes of their trading partners and warn of potential credit risks.

[0425] (e) Calibration of predictability through regression learning of decision paths The system can acquire the user's decision-making process as an attribute and evaluate and update predictability based on its consistency with the results of the time delay. This allows experiences, including failures, to be accumulated as learning resources, contributing to the improvement of long-term decision-making capabilities.

[0426] (f) Providing distillation know-how and metacognition It can extract highly predictable decision types from the decision-making processes of a large number of users and display the distance from those decisions to user decisions. Furthermore, it enables consistency comparisons with populations defined by time, industry, region, etc., presentation of time-series changes in similar samples, and presentation of statistical or probabilistic results.

[0427] (g) Construction of an external observation model involving a third party For third-party entities other than the user, an external observer model 104 can be generated to identify internal responses from externally observable inputs and outputs (see Figure 2). This improves the accuracy of predicting the behavior of trading partners, competitors, potential partners, etc.

[0428] (h) Improving the accuracy of external instrument models through sequential search and collection By dynamically searching for and collecting information about the target entity, the predictive performance of the external observation instrument model 104 can be continuously improved.

[0429] (i) Implementing focus on specific targets as a load allocation option. Under limited resources, it is possible to establish pathways that enhance the learning intensity for specific targets in accordance with the strategic needs of the user.

[0430] (j) Achieving both individual optimization and overall optimization By redirecting additional resources towards improving overall performance, it becomes possible to simultaneously address the challenges faced by individual users and enhance the predictive capabilities of the entire information processing method.

[0431] (k) Ensuring transparency in evaluation criteria By explicitly allowing the calibration conditions to be switched, it is always possible to clearly understand "what assumptions" the evaluation results are based on.

[0432] (l) Supporting strategic decision-making By virtually changing the attributes of the evaluation entity, it is possible to support strategic decisions such as "For whom is this intangible asset 200 most valuable?" and "To whom should it be sold?"

[0433] (m) Ensuring metrological rigor (integration effect with Example 9) In this embodiment, the concept of measurement resolution 401 introduced in Embodiment 9 is retrospectively applied to the distance evaluation (value deviation (Δ) 411, norm deviation, judgment path distance, etc.), thereby suppressing overinterpretation of apparent slight differences and improving the validity of the measurement (see Figure 8).

[0434] (n) Substantive evaluation of international law, treaties, etc. (unique effects of this embodiment) Regarding international law, treaties, and statements, it becomes possible to evaluate them not based on formal legal interpretation, but on substantive effects such as the process of agreement formation, legitimacy, discrepancies in interpretation, and performance. This makes the discrepancy between form and substance visible, allowing for the identification of true credibility or risk.

[0435] In particular, by using the degree of deviation from the wording of international law and other regulations, as well as time-series stability, as the main indicators, and predicting the likelihood of future compliance, the accuracy of decision-making in international transactions, investment decisions, contract strategies, etc., will be dramatically improved.

[0436] The information processing method according to this embodiment improves the validity of measurements in the evaluation of intangible assets 200 and rights, as well as in decision support, by adding new technical value such as calibration of evaluation criteria, regression learning of decision paths, dynamic construction of external observer model 104 (see Figure 2), and evaluation of the substantive effects of international law and treaties, on top of the technical foundation established in Examples 1 to 3. [Examples]

[0437] (Examples of disclosure control, transparency assurance, and auditing functions for decision-making path attributes)

[0438] The following describes yet another embodiment of the information processing method according to the present invention.

[0439] (0) Positioning of this embodiment and relationships between embodiments This embodiment is constructed based on the following embodiment.

[0440] Direct premise (required): In Example 4, regression learning and predictability evaluation of the decision path, identification of the external observer model 104, and distance evaluation with the reference module 410 are introduced (see Figures 2 and 9).

[0441] Examples 1 to 3 detail the definition of the asset elements (201 to 205) of the intangible asset 200, the introduction of the latent space 210, the projection 222 onto the observation space 220, and the observation function of time series transitions (see Figures 3 and 5).

[0442] Indirect premise (foundation): The metaphors such as "boundary," "fortress wall," and "insect infestation" used in Example 2 are applied as means to represent the audit scope, audit confidence level of 601, and tampering risk in this example (see Figure 6).

[0443] Uniqueness of this embodiment: In this embodiment, we provide a technical means that reconciles the conflicting demands of privacy protection through anonymization and trust building through selective disclosure of sensitive information such as decision-making path attributes, through explicit selection by the user. Furthermore, it is unique in that it integrates a social verification means, such as third-party audits, into the information processing method and quantifies and visualizes the value of the audit as a distance (see Figure 11).

[0444] Effects on subsequent embodiments: The concepts of audit record 600 and confidence level 601 introduced in this embodiment are used in the confidence evaluation in the dynamic learning of the external observer model 104 in Embodiment 6, applied to the calculation of stratified confidence levels in the multilayer evaluation model in Embodiment 7, and used for confidence level display in decision support in Embodiment 8 (see Figure 11).

[0445] This embodiment is based on the regression learning and predictability evaluation of the decision path described in Example 4, and includes configurations for disclosing and controlling the attributes of the decision path, ensuring transparency, and third-party audits (see Figures 2 and 11).

[0446] In this embodiment, the system provides a function that allows users to strategically control the scope of disclosure of their own decision-making pathway attributes, while balancing privacy protection and trust building, and to undergo audits by third parties as needed, with the results made visible.

[0447] (1) The need to control the disclosure of judgment path attributes

[0448] As explained in Example 4, the decision path (operation sequence, selection history, weighting, etc.) is important attribute information that represents the user's decision-making characteristics.

[0449] The predictability score calculated from this decision-making process can function as an indicator of user reliability, but at the same time, this information carries the risk of revealing the user's strategies, thinking patterns, weaknesses, etc.

[0450] Therefore, in handling the attributes of the decision-making process, it is necessary to reconcile the following conflicting requirements.

[0451] (Request A) Privacy protection Protecting information that could identify individual users, or information that should be strategically kept confidential.

[0452] (Request B) Building trust Gaining trust from trading partners, investors, potential partners, etc., by demonstrating one's predictiveness score or judgment ability.

[0453] This embodiment technically implements these conflicting requirements as an explicit option for user-controlled disclosure.

[0454] (2) Anonymization and concealment of decision-making path attributes (default setting)

[0455] (2-1) Anonymization process as a general principle In the information processing method according to this embodiment, the learning of decision path attributes is performed, in principle, after going through the following processes.

[0456] (a) Anonymization Remove or pseudonymize identifiers that can identify individuals or organizations (such as user IDs, organization names, and IP addresses).

[0457] (b) Aggregation Individual decision paths are aggregated with multiple paths that have similar patterns to reduce individuality.

[0458] (c) Confidentiality We apply differential privacy, k-anonymity, or similar concealment techniques to reduce the risk of re-identifying individual users.

[0459] As a result, the extraction of distillation know-how and the provision of metacognition in Example 4 are performed without infringing on the privacy of individual users.

[0460] (2-2) Adjustment of the degree of secrecy The degree of anonymization (for example, the ε value in differential privacy, the k value in k-anonymity) may be configurable by the user.

[0461] Increasing the degree of anonymization (decreasing ε, increasing k) enhances privacy protection, but may decrease the accuracy of the predictability score.

[0462] Reducing the degree of anonymization may improve the accuracy of the predictability score, but it could also increase the risk of re-identification.

[0463] This information processing method may present the trade-off to the user and set the anonymization parameters based on the user's choice.

[0464] (3) Selective disclosure of judgment path attributes

[0465] (3-1) Motivation and purpose of disclosure On the other hand, individuals or organizations may want to make their own decision-making pathways more transparent in order to improve the level of trust others place in them (see Figure 11).

[0466] For example, the following situations are conceivable: when you want to demonstrate your investment judgment ability to investors; when you want to show your credibility to trading partners; when you want to appeal your strategic thinking ability to potential partners; and when you want to demonstrate your compliance to regulatory authorities.

[0467] Based on these motivations, users may choose to disclose some or all of their decision-making pathway attributes to specific parties.

[0468] (3-2) Setting the scope of disclosure The information processing method according to this embodiment may provide the user with a function to set the range of decision path attributes to be disclosed.

[0469] The scope of disclosure may be defined, for example, by the following dimensions:

[0470] (a) Types of attributes Which attributes should be disclosed from among viewing order, time spent, weighting operations, and final judgment?

[0471] (b) Time range Which period of the past should the decision-making process be disclosed for (e.g., the past 6 months, the past 3 years, or the entire period)?

[0472] (c) Target area Which business areas, markets, or evaluation categories will have their decision-making processes disclosed?

[0473] (d) Aggregation level Should we disclose individual decision-making processes, or should we only disclose aggregated statistical values ​​(mean, variance, success rate, etc.)?

[0474] (3-3) Designation of recipient of disclosure Furthermore, the information processing method according to this embodiment may also provide a function for specifying the recipient of disclosure according to the following criteria.

[0475] (a) Individual designation A specific individual or organization is designated by an identifier.

[0476] (b) Attribute specification The recipients of the disclosure are defined based on attributes such as industry, region, company size, and credit score.

[0477] (c) Conditional disclosure Disclosure will only be made if the other party meets specific conditions (e.g., mutual disclosure, conclusion of a confidentiality agreement, presentation of audit certificates, etc.).

[0478] (3-4) Separation of learning contributions on a per-user basis The information processing method according to this embodiment may include a configuration that allows the user to choose whether or not to contribute the judgment path attributes, evaluation inputs, or other information provided by the user to the learning of the entire information processing method.

[0479] If a user chooses not to allow contributions, their information will only be used for their own specific calibration and predictions and will not contribute to improving the accuracy of other users' evaluations.

[0480] This configuration achieves both the protection of trade secrets and other confidential information, and the improvement of the overall evaluation accuracy of this information processing method.

[0481] (4) Third-party audit function

[0482] (4-1) Purpose and significance of audits When disclosing decision-making pathway attributes, it is desirable that third parties be able to verify that these attributes are accurate, have not been tampered with, and that the claimed predictability score has been properly calculated.

[0483] In this embodiment, a third-party audit function for decision path attributes and predictability scores is provided (see Figure 11).

[0484] The aforementioned audit has the following objectives: to verify the accuracy of the disclosed decision path attributes, to verify the validity of the predictability score calculation process, to detect falsification, fabrication, or selective disclosure (presentation of only favorable data), and to record the audit results and preserve the evidence.

[0485] (4-2) The entity that conducts the audit The aforementioned audit may be conducted by any of the following entities:

[0486] (a) Auditing body with legal backing Certified public accountants, audit firms, certified information security auditors, and other entities authorized to audit by law or regulation.

[0487] (b) Independent third-party organization Industry associations, non-profit organizations, or similar independent bodies.

[0488] (c) Technical Verification Services A specialized company that provides technical tools such as blockchain auditing and cryptographic verification.

[0489] (d) Self-audit Internal audits conducted by the users themselves, however, have limited independence.

[0490] This information processing method identifies the type of auditing entity and reflects its confidence level of 601 in the evaluation (see Figure 11).

[0491] (4-3) Structure of audit records In this embodiment, the audit record 600 may include the following information (see Figure 11).

[0492] (a) Audit entity identifier 602 Identification information of the entity that conducted the audit (name, authentication ID, public key, etc.).

[0493] (b) Audit date 603 The date and time (timestamp) when the audit was conducted.

[0494] (c) Identifiers of audited features An identifier that specifies the scope, period, type, etc., of the decision-making path attributes that were subject to the audit.

[0495] (d) Audit results Audit opinions such as "acceptable," "conditionally acceptable," "unacceptable," or "no opinion," as well as facts or findings discovered during the audit process.

[0496] (e) Signature information 604 for tamper detection A digital signature, hash value, or transaction identifier on the blockchain to detect tampering with the audit record 600 itself.

[0497] (4-4) Disclosure of audit records The aforementioned audit record 600 may be disclosed simultaneously with the disclosure settings for the judgment path attributes.

[0498] In other words, when a user discloses decision-making path attributes to a specific party, the corresponding audit record 600 is also disclosed, thereby ensuring the reliability of the disclosed information (see Figure 11).

[0499] (5) Assessment and visualization of audit value

[0500] (5-1) Definition of the audit value vector The information processing method according to this embodiment may define an audit value vector in order to quantitatively evaluate the value of the audit.

[0501] The aforementioned audit value vector may include, for example, the following components:

[0502] (a) Auditor confidence score A score based on factors such as legal backing, degree of independence, past audit performance, and industry reputation.

[0503] (b) Comprehensiveness of the audit scope To what extent are the audited decision-making pathway attributes comprehensive to the entire scope of disclosure?

[0504] (c) Audit depth Was the verification limited to a superficial check, or was a detailed verification (sampling, recalculation, falsification search, etc.) conducted?

[0505] (d) Novelty of the audit The elapsed time from audit date 603 to the present (the older the audit, the lower the confidence level of 601).

[0506] (e) Strength of tamper detection Tamper resistance based on the signature technology used, hash algorithm, and the decentralized nature of the blockchain.

[0507] (5-2) Distance from the standard audit vector This information processing method may also calculate the distance between the audit value vector and a predetermined standard audit vector.

[0508] The aforementioned standard audit vector, similar to standard module 410 in Example 4, represents a normative audit level and may be defined based on, for example, the following (see Figure 9): recommended audit levels in industry guidelines, audit standards required by regulatory authorities, international audit standards (ISA, etc.), and audit requirements independently set by the user.

[0509] If the aforementioned distance is small, the audit is evaluated as conforming to the standards; if the distance is large, the reliability of the audit is evaluated as limited.

[0510] (5-3) Visualization of audit value In this embodiment, the information processing method may visualize the distance between the audit value vector and the reference audit vector on the observation space 220 (see Figures 6 and 11).

[0511] Specifically, the following display methods are possible.

[0512] The confidence level of the audit (601) is displayed using color (e.g., high confidence = green, medium confidence = yellow, low confidence = red). The comprehensiveness of the audit scope is represented as the continuity or thickness of the boundary in the "boundary" and "wall" metaphors of Example 2. The recency of the audit is represented as transparency or brightness (older audits are displayed more faintly). The strength of the tamper detection is represented as the difficulty of the intrusion route in the "insect infestation" metaphor of Example 2.

[0513] (6) Treatment of value differences by the auditing body

[0514] (6-1) Differences in confidence due to differences in auditing bodies In this embodiment, the information processing method explicitly handles the fact that the confidence level of the audit (601) differs depending on the type of auditing entity (whether or not there is legal backing, the degree of independence, etc.) (see Figure 11).

[0515] For example, the following differences in evaluation are possible.

[0516] (a) Audit by a legally compliant auditing body Audits conducted by certified public accountants, audit firms, etc., are assigned a high confidence score because independence, confidentiality, and professional ethics are guaranteed in accordance with the law.

[0517] (b) Audit by an independent third-party organization Audits conducted by industry associations and similar bodies have a certain degree of independence, but their legal enforceability is limited, resulting in them being assigned a moderate confidence score.

[0518] (c) Audit by technical verification services While cryptographic verification offers strong tamper detection capabilities, its confidence score is only partial when the audit scope is limited to technical aspects.

[0519] (d) Self-audit Audits conducted by users themselves lack independence and therefore receive the lowest confidence score.

[0520] (6-2) Representation as distance The aforementioned confidence difference is expressed by the concept of "distance" introduced in Example 4 (see Figure 9).

[0521] In other words, the distance between the audit value vector and the standard audit vector changes depending on the type of auditing entity, and this distance is reflected in the presentation.

[0522] This allows users to visually understand the extent to which the disclosed decision path attributes and predictability scores are based on reliable audits (see Figure 11).

[0523] (6-3) Handling of bias in external modules and information processing models In this embodiment, the same concept as the confidence difference of the auditing entity in (6-1) and (6-2) above is also applied to the external module, the auditing entity, and the information processing model (e.g., LLM) used in this information processing method.

[0524] It is self-evident that when these entities or models make evaluations or judgments, they themselves possess inherent biases (such as biases stemming from training data, design assumptions, and limitations on their scope of application).

[0525] Therefore, this information processing method may include a function to present the following information to the user.

[0526] (a) Distance from the reference point of the external module For the external module used in Example 3, the evaluation characteristics of the external module (scope of application, area of ​​expertise, known bias tendencies, etc.) and the distance from a predetermined reference module 410 are calculated and displayed to the user (see Figure 9).

[0527] (b) Distance from the reference of the information processing model (LLM, etc.) For the information processing models (e.g., LLM) used in Examples 2 and 8, the distance between the known bias characteristics of the model (such as bias in the training data, tendency to overestimate / underestimate in specific domains, and biases due to language and culture) and a neutral standard is estimated and displayed to the user.

[0528] (c) Distance from the auditing body's standards For the auditing entities described in (6-1) and (6-2) above, the distance between the auditing entity's evaluation tendencies and normative auditing standards is shown.

[0529] The aforementioned distance may be expressed using the concept of "distance" introduced in Example 4 and visualized on the observation space 220 (see Figure 6).

[0530] This allows users to interpret the evaluation results and make appropriate decisions, after recognizing the potential biases in the external modules, information processing models, and auditing entities on which this information processing method relies.

[0531] Therefore, this configuration has the effect of improving the transparency of evaluation results, reducing the risk of users over-relying on evaluation results, and supporting autonomous decision-making.

[0532] (6-4) Metaphorical expression through facial expressions In the information processing method according to this embodiment, when displaying the confidence level 601 or the degree of deviation from the standard, an expression may be added to the object (including metaphors) placed on the observation space 220 (see Figures 6 and 11).

[0533] Prior art has shown that switching the facial expressions (demeanors) of an avatar can influence the trustworthiness perceived by the user (see Patent Document 5). Furthermore, techniques are known for generating facial expression parameters from text, audio, or video, and rendering avatars based on eye movements, mouth movements, head posture, etc. (see Patent Document 6).

[0534] In this embodiment, these prior art techniques are applied to represent the confidence level 601, audit value, or deviation from the standard in the valuation of intangible assets 200 as a metaphorical expression (see Figure 3).

[0535] Specifically, for the metaphors such as insect damage exemplified in Example 2, or other character representations, facial expressions are assigned according to a confidence level of 601 or a degree of deviation. Here, facial expressions are visual features composed of facial parameters such as the angle of the eyes, the angle of the mouth, the position of the eyebrows, the tilt of the entire face, and the direction of the gaze.

[0536] When expressing the differences in the reliability of auditing entities as described in (6-1) above using facial expressions, high reliability (audits by auditing entities with legal backing, etc.) may be represented by facial expressions that give a trustworthy impression (for example, a forward gaze, gentle eyes, upturned corners of the mouth, etc.), while low reliability (self-audits, etc.) may be represented by facial expressions that give an untrustworthy impression (for example, averted gaze, downturned corners of the mouth, asymmetrical eyebrows, etc.).

[0537] Similarly, the degree of bias in the external module or information processing model described in (6-3) above may be expressed as a metaphorical facial expression symbolizing the module. If the bias is large, an expression that suggests a deviation from a neutral standard (e.g., a tilted head, asymmetrical mouth, etc.) can be assigned, allowing users to intuitively receive a warning when interpreting the evaluation results of the module.

[0538] In particular, in displaying behavioral differences based on counterparty attributes detected in Example 4 (3-4), facial expressions are effective. That is, when an entity exhibits different behaviors depending on the attributes of its trading partner (e.g., strong / weak party, transaction size, etc.) (application of double standards), by assigning the metaphor of "untrustworthy facial expression" to this non-uniformity of behavior, users can intuitively recognize the credit risk in transactions with that entity.

[0539] For example, in the case of an entity that refrains from exercising its rights against large corporations while actively exercising its rights against small and medium-sized enterprises, the existence of a double standard may be visually suggested by assigning a metaphor representing that entity an expression with an unfocused gaze or an expression that gives different impressions on the surface and underneath.

[0540] In generating the aforementioned facial expression parameters, a method may be used that extracts angle information of the eyes and mouth from the landmark coordinates of the face and applies this angle information to the avatar or character representation, similar to the technique disclosed in Patent Document 6. Alternatively, a mapping function or a pre-trained model may be used that takes a confidence score or deviation as input and outputs the corresponding facial expression parameters.

[0541] These facial expression parameters are applied to a two-dimensional character image, a three-dimensional avatar model, or an animation representation, and are rendered on the observation space 220 (see Figure 6).

[0542] This configuration allows this information processing method to convert abstract numerical indicators (confidence score, deviation degree, bias level, etc.) into visual representations of facial expressions that humans can instinctively and immediately understand. As a result, users can grasp an overview of the reliability or credit risk of the subject simply by glancing at the metaphorical facial expression, without having to refer to numerical values ​​or graphs in detail, thereby reducing cognitive load and accelerating decision-making.

[0543] In this embodiment, the display using facial expressions may be used in combination with the display using color, transparency, brightness, boundary continuity, etc., as exemplified in (5-3) above, or it may be selectively switched on or off depending on the user's settings.

[0544] (7) Clarification of the relationships between the embodiments This embodiment is built upon the concept of Embodiment 4, as described below.

[0545] Relationship with Example 4: The "regression learning of decision paths" introduced in Example 4 is extended in this embodiment as a disclosure control function to balance privacy protection and transparency. Furthermore, the "predictability score" in Example 4 is positioned in this embodiment as an index 207 that can be verified by a third-party audit. In addition, the concept of the external observer model 104 in Example 4 is applied in this embodiment to the evaluation of the auditing entity (see Figure 2).

[0546] Relationship with Examples 1-3: The concepts of the latent space 210, the projection 222 onto the observation space 220, and distance evaluation constructed in Examples 1 to 3 are used in this embodiment as means of visualizing audit value (see Figures 5 and 6). In particular, the metaphors of "boundary," "wall," and "insect infestation" in Example 2 are applied in this embodiment as means of representing the audit scope, audit confidence 601, and tampering risk.

[0547] Uniqueness of this embodiment: In this embodiment, we provide a technical means that reconciles the conflicting demands of privacy protection through anonymization and trust building through selective disclosure of sensitive information such as decision-making path attributes, through explicit selection by the user. Furthermore, it is unique in that it integrates a social verification means, such as third-party audits, into the information processing method and quantifies and visualizes the value of the audit as a distance (see Figure 11).

[0548] (8) Effects As described above, the following effects can be obtained according to this embodiment.

[0549] (a) Ensuring the protection of privacy As a general rule, anonymization, aggregation, and confidentiality are implemented to protect user privacy in the learning of decision-making pathway attributes.

[0550] (b) Achieving strategic transparency When a user wishes to prove their own decision-making capacity, they can selectively disclose their decision-making process attributes while controlling the scope and recipients of the disclosure.

[0551] (c) Support for building trust By presenting decision-making pathway attributes and predictability scores verified by third-party audits, companies can gain the trust of trading partners, investors, potential partners, and others (see Figure 11).

[0552] (d) Visualization of audit reliability The audit value, based on the type of auditing entity, audit scope, audit depth, etc., can be quantified as a distance and presented visually (see Figure 11).

[0553] (e) Realization of tamper detection Tampering with audit records 600 can be detected using technical means such as digital signatures, hash values, and blockchain.

[0554] (f) Facilitating regulatory compliance By setting the audit standards required by regulatory authorities as the benchmark audit vector, compliance can be demonstrated.

[0555] (g) Designing incentives for disclosure Users who undergo high-reliability audits receive a higher rating for the reliability of their predictability score, creating an incentive to undergo audits.

[0556] (h) Mitigation of information asymmetry Because users with high judgment abilities can demonstrate those abilities in a verifiable way, adverse selection (where only users with low abilities remain in the market) can be prevented.

[0557] (i) Transparency of bias (unique effect of this embodiment) By visualizing biases inherent in external modules, information processing models (such as LLMs), and auditing entities as distances from the standard, users can appropriately interpret evaluation results and make autonomous decisions.

[0558] (j) Intuitive understanding through facial expressions (a unique effect of this embodiment) By representing a confidence score of 601 or the degree of bias as a metaphorical facial expression, users can intuitively grasp the reliability of the subject without having to refer to the numerical values ​​in detail (see Figure 11).

[0559] The information processing method according to this embodiment improves the reliability of valuing intangible assets 200 and rights by adding social, ethical, and technical values ​​such as privacy protection, transparency assurance, and third-party verification to the regression learning function of the decision path constructed in Embodiment 4 (see Figures 3 and 11). [Examples]

[0560] (Examples of acquiring, harmonizing, placing into potential space and applying to the calibration of valuation criteria for publicly available intangible asset information, as well as rights management and consideration distribution)

[0561] The following describes yet another embodiment of the information processing method according to the present invention.

[0562] This embodiment assumes the calculation method for intangible assets 200 and indicators 207 related to rights described in Example 1 (see Figure 3), the observation of time-series transitions described in Example 2, the control of projection 222 from latent space 210 to observation space 220 described in Example 3 (see Figure 5), and the calibration of evaluations based on evaluation subject attributes and normative criteria described in Example 4 (see Figure 9), and shows a configuration for integrating publicly available intangible asset information into the evaluation system of this information processing method, as well as the mechanisms for rights processing, trust management, and consideration distribution in the use of said information.

[0563] In this embodiment, the concept of calibration introduced in Embodiment 4 is extended to provide a technical means for handling not only the intangible assets 200 directly owned by the user, but also knowledge about intangible assets contained in publicly available literature, case studies, or trained models, on a unified evaluation platform, while also ensuring the protection of rights holders and appropriate compensation in the provision of such knowledge (see Figures 1 and 2).

[0564] (1) Definitions of terms and relationships between examples

[0565] (1-1) Definition of publicly available intangible asset information In this embodiment, "publicly available intangible asset information" refers to information that falls under any of the following categories:

[0566] This includes descriptions of intangible assets found in literature such as books, papers, reports, and research materials; autobiographies, interview transcripts, and oral accounts by managers and experts; case studies on the creation, growth, damage, or extinction of intangible assets; structured datasets (such as patent databases and corporate financial information); and feature representations contained within trained models (e.g., LLMs) that have learned from the aforementioned information.

[0567] The aforementioned information includes not only success stories, but also failure stories, partially successful stories, or stories involving multiple contributing factors.

[0568] The aforementioned publicly available intangible asset information is reference knowledge used for evaluation and differs in nature from the external environmental information (information indicating the external conditions of the subject of evaluation) mentioned in Example 2. The aforementioned publicly available intangible asset information is obtained from the external information source 30 (see Figure 1).

[0569] (1-2) Relationship with Examples 1-5 This embodiment is connected to the existing embodiment as follows.

[0570] Relationship with Example 1: The asset elements (elements 1-5: symbols 201-205) defined in Example 1 are used as classification criteria for publicly available information in this embodiment (see Figure 3). The asset elements extracted from publicly available information are classified according to the definition in Example 1 and represented as combination state 206.

[0571] Relationship with Example 2: The concept of standard documentation introduced in Example 2 is extended in this embodiment to serve as a reference base for evaluating publicly available intangible asset information. Furthermore, the time-series transition observation function in Example 2 is used to reproduce past state changes contained in the publicly available information on the latent space 210 (see Figure 5). The external observer function in Example 2 forms the basis for the function of utilizing publicly available information as a reference base in this embodiment.

[0572] Relationship with Example 3: The latent space 210 detailed in Example 3 functions in this embodiment as a space for arranging the indicators 207 extracted from publicly available information (see Figure 5). Furthermore, the concept of similarity rate in Example 3 is applied when evaluating the similarity between publicly available examples and the current state of the user.

[0573] Relationship with Example 4: The calibration concept introduced in Example 4 is central to this embodiment (see Figure 9). Publicly available information is treated as a set of historical documents constituting the external observer function in Example 4 and is used for calculating the intrinsic value 413 of the evaluation subject and for evaluating the distance with the reference module 410. The external observer model 104 in Example 4 forms the basis of the function that dynamically incorporates publicly available information as training data in this embodiment (see Figure 2).

[0574] Relationship with Example 5: The concept of disclosure control detailed in Example 5 is applied in this embodiment to control the scope of use and display granularity of the publicly available information.

[0575] (2) Acquisition of publicly available intangible asset information

[0576] (2-1) Information source This information processing method may include a step of obtaining publicly available intangible asset information from the following entities via an external information source 30 (see Figure 1).

[0577] (a) Provided by the user The user provides their own literature, materials, or knowledge as input through the input unit 22.

[0578] (b) Provision by a third party Provision of information created by researchers, experts, data providers, etc.

[0579] (c) Provision by the provider of this information processing method The platform operator provides information that has been collected and compiled.

[0580] (d) Automatic acquisition Automatic acquisition via RPA or API integration as mentioned in Example 2 (see Figures 1 and 10).

[0581] (2-2) Format of information The aforementioned information may be in any of the following formats:

[0582] This includes descriptions in natural language (books, papers, reports, etc.), structured data (databases, spreadsheets, etc.), semi-structured data (XML, JSON, etc.), and internal representations of trained models (embedding vectors, features, etc.).

[0583] (3) Structure of the rights management organization

[0584] (3-1) Necessity of rights management When the information processing method according to this embodiment presents publicly available intangible asset information to the user by quoting or referring to it in the explanation request function or the presentation of branching reasons for future state 300 in Embodiment 3, the following rights may be involved with such information (see Figure 7).

[0585] (a) Copyright Copyrights relating to written expressions, diagrams, photographs, videos, etc., in books, papers, reports, etc.

[0586] (b) neighboring rights Rights of performers, record producers, broadcasters, etc.

[0587] (c) Rights under the Unfair Competition Prevention Act Rights relating to trade secrets, limited-access data, product labels, etc.

[0588] (d) Publicity rights The right to commercially use the names, likenesses, etc., of famous people.

[0589] (e) portrait rights The right not to have one's appearance or figure photographed or published without permission.

[0590] (f) Other intellectual property rights Intellectual property rights, such as design rights and trademark rights, that are included in publicly available information.

[0591] In this embodiment, we provide a technical means to appropriately handle the aforementioned rights, protect the interests of the rights holders, and realize effective information provision to users.

[0592] (3-2) Acquisition of rights information and assignment to metadata This information processing method may include a step of acquiring rights information related to publicly available intangible asset information when acquiring such information.

[0593] The aforementioned rights information may include the following: rights holder identification information (information identifying authors, copyright holders, performers, and other rights holders), type of right (copyright, neighboring rights, publicity rights, portrait rights, and other types of rights), scope of right (specific rights such as the right of reproduction, the right of adaptation, and the right of public transmission), license terms (the scope of use and the conditions of use), validity period of the right (copyright protection period, contractual license period, etc.), and whether or not a trust exists (the status of the conclusion of a rights trust agreement, as described later).

[0594] The aforementioned rights information is attached as metadata to publicly available intangible asset information and is managed in conjunction with confidence level 601, update date, etc.

[0595] (3-3) Determination of compliance with patent rights The information processing method according to this embodiment may include a step of determining, based on the rights information, whether the presentation of publicly available intangible asset information to a user constitutes an infringement of rights.

[0596] The aforementioned determination may include the following processes.

[0597] (a) Determining whether a quotation is appropriate Determining whether the requirements of Article 32 (Quotation) of the Copyright Act are met. Specifically, this involves confirming that the work is a published work, conforms to fair practice, is within a reasonable scope for the purpose of reporting, criticism, research, or other purposes of quotation, that the relationship between the quoted portion and the main text is clear, and that the source is clearly indicated.

[0598] (b) Limitation on the number of quoted characters Limit the number of characters quoted from a single source to a predetermined maximum (e.g., less than 15 words).

[0599] (c) Performing summarization and extraction processing If displaying the full text may infringe on rights, the work will be converted into a format that does not reproduce the essential parts of the work through summarization, keyword extraction, or structured extraction.

[0600] (d) Confirmation of publicity rights and portrait rights If the publicly available information includes the names, likenesses, etc., of famous people, we will confirm that the use of such information is not solely for the purpose of exploiting the customer-attracting power of those famous people.

[0601] (e) Confirmation of whether it constitutes a trade secret. We will confirm that the publicly available information does not constitute a trade secret under the Unfair Competition Prevention Act. We will refrain from disclosing information that may potentially constitute a trade secret.

[0602] If the determination process determines that the risk of infringement exceeds a predetermined threshold, the information processing method may suppress the presentation of the information or display a warning to the user on the display unit 21 (see Figure 1).

[0603] (3-4) Implementation of rights-compliant information presentation The information processing method according to this embodiment may include a step of presenting information in one of the following ways based on the result of the patent compliance determination.

[0604] (a) Lawful methods of citation Quoting in a form that meets the requirements of Article 32 of the Copyright Act. This includes ensuring clear indication of the source, clarification of the quoted portion, and maintenance of the principal-subordinate relationship.

[0605] (b) Summary presentation method The original text is summarized and presented in a format that does not reproduce the essential parts of the copyrighted work. The summary is generated by this information processing method or an external summarization model.

[0606] (c) Structured information presentation method This structured data presents only factual information (dates, amounts, names of people, places, etc.) extracted from publicly available sources. It does not include creative expressions from copyrighted works.

[0607] (d) Link reference method Instead of directly presenting the content of publicly available information, only the location of the information (URL, bibliographic information, etc.) will be provided, allowing users to refer to the original source.

[0608] (e) Full text presentation method for entrusted information Based on the rights trust agreement described later, we will present the full text or a detailed extract of the information for which we have received permission to use from the rights holder.

[0609] The selection of the presentation method may be automatically determined based on rights information, license terms, and user settings (such as requirements for level of detail), or it may be manually selected by the user from the input unit 22 (see Figure 1).

[0610] (4) Rights trust agreement and trust management organization

[0611] (4-1) Definition of a rights trust agreement In this embodiment, "rights trust agreement" means an agreement in which a rights holder (copyright holder, publicity rights holder, etc.) relating to publicly available intangible asset information entrusts the management of said rights to the provider of the information processing method or a designated trust administrator.

[0612] The aforementioned rights trust agreement may include the following clauses: the purpose of the trust (that the rights holder's copyrighted works, portraits, and other intellectual property will contribute to the development of industry in society, the sharing of knowledge, the valuation of failure cases, etc., by being provided to users through this information processing method); the trust property (the scope and content of the rights subject to the trust); the scope of the license (the form of use, purpose of use, and scope of users in this information processing method); the method of calculating the consideration (the method of calculating the consideration based on actual use); the method of distributing the consideration (the method of distributing the calculated consideration to the rights holder, the distribution cycle, the minimum payment amount, etc.); the trust fee (the method of calculating the trust fee received by the trust administrator); and the conditions for termination of the contract (the conditions for termination of the contract by the rights holder or the trust administrator).

[0613] The aforementioned rights trust agreement may be concluded individually between the rights holder and the trust administrator, or it may be presented as standard contract terms and conditions and established upon the consent of the rights holder.

[0614] (4-2) Functions of the trust administrator In this embodiment, the trust administrator has the following functions.

[0615] (a) Registration and management of rights holders This system registers and manages the rights holder's identification information, contact details, rights details, and license terms.

[0616] (b) Rights clearance for publicly available information Regarding the publicly available intangible asset information provided through this information processing method, the permission status of the rights holder will be confirmed and registered as usable information.

[0617] (c) Records of actual usage The actual usage of publicly available information in this information processing method (viewing, citation, evaluation, etc.) is recorded as update history 208 (see Figure 3).

[0618] (d) Calculation and distribution of consideration The compensation will be calculated based on actual usage and distributed to the rights holders.

[0619] (e) Reporting to rights holders The rights holders will be regularly informed of the actual usage, the basis for calculating the compensation, and the amount of distribution.

[0620] The aforementioned trust administrator may be the same as the provider of this information processing method, or it may be established as an independent third-party organization.

[0621] (4-3) Management of rights trust information The information processing method according to this embodiment may include a step of managing information based on a rights trust agreement in the storage unit 12 as metadata for publicly available intangible asset information (see Figure 1).

[0622] The aforementioned rights trust information may include the following: a trust flag (a flag indicating whether or not the information is subject to a rights trust agreement), rights holder identification information (identification information of the rights holder who entered into the trust agreement), usage level (level of permitted usage, such as summary only, extraction only, full text display allowed, etc.), consideration calculation method (consideration calculation method such as viewing fee, citation fee, fixed fee, etc.), and contract expiration date (expiration date of the rights trust agreement).

[0623] The aforementioned rights trust information is treated as subject to disclosure control in Example 5 and is not disclosed to any entity other than the rights holder.

[0624] (5) Recording of actual usage and calculation of fees

[0625] (5-1) Records of actual usage The information processing method according to this embodiment may include a step of recording the usage status as an update history 208 when a user views, quotes, or evaluates publicly available intangible asset information (see Figure 3).

[0626] The record of usage may include the following: date and time of use (date and time the information was used, timestamp), user identification information (user identification information, anonymized ID, etc.), form of use (form of use such as viewing, quoting, summarizing, displaying the full text, inputting evaluations, etc.), amount of use (amount of use such as viewing time, number of quoted characters, number of displays, etc.), purpose of use (purpose of use such as evaluation, comparison, explanation, learning, etc., which can be specified by the user), and information identifier (identifier of the publicly available intangible asset information used).

[0627] The records of the aforementioned usage patterns are treated as part of the decision-making path attributes in Example 5, and personally identifiable information is anonymized or aggregated to protect privacy.

[0628] (5-2) Calculation of consideration The information processing method according to this embodiment may include a step of calculating the consideration to be paid to the rights holder based on the record of actual usage.

[0629] The aforementioned consideration may be calculated by any of the following methods:

[0630] (a) Per-use pricing method The cost is calculated by multiplying the number of uses (such as views and citations) by the unit price.

[0631] (b) Time-based pricing The calculation is done by multiplying the viewing time, usage time, etc., by the unit price.

[0632] (c) Price per character The calculation is done by multiplying the number of quoted characters by the unit price.

[0633] (d) Evaluation-linked method Based on user evaluations (star ratings, usefulness ratings, etc.), information that receives a high rating will be compensated with a higher price.

[0634] (e) Fixed amount method A fixed fee will be calculated at regular intervals, regardless of actual usage.

[0635] (f) Combination method The calculation is performed by combining several of the above (a) to (e).

[0636] The aforementioned calculation method is determined in advance in the rights trust agreement.

[0637] (5-3) Aggregation and distribution of consideration This information processing method may include a step of aggregating the compensation for each rights holder at predetermined intervals (e.g., monthly, quarterly, etc.) and distributing it to the rights holders.

[0638] The aforementioned distribution may be carried out in accordance with the following procedure: setting the aggregation period (setting the period for aggregating the consideration), calculating the consideration for each rights holder (calculating the consideration for each rights holder based on the calculation method in (5-2) above), deducting the trust fee (deducting the trust fee received by the trust administrator from the calculated consideration), confirming the minimum payment amount (confirming whether the consideration after deduction exceeds the contractual minimum payment amount. If it falls below the minimum payment amount, it will be carried over to the next period), executing the payment (executing the payment to the rights holder by bank transfer to the designated account, electronic payment, etc.), and sending the payment details (sending the rights holder a payment details that include details of the actual usage, the basis for calculating the consideration, the breakdown of the trust fee, etc.).

[0639] The distribution process is recorded in a format that allows for tamper detection using a timestamp, digital signature, etc., similar to the audit record 600 in Example 5 (see Figure 11).

[0640] (6) Priority presentation of information that has been placed in trust

[0641] (6-1) Setting the presentation priority In this embodiment, the information processing method may set a higher presentation priority for publicly available intangible asset information based on a rights trust agreement compared to information that is not in trust.

[0642] The following effects can be obtained by setting the aforementioned presentation priority.

[0643] (a) Improving the value provided to users Since the entrusted information can be displayed in full or extracted in detail, it becomes possible to provide users with more detailed information.

[0644] (b) Facilitating the return of compensation to rights holders Prioritizing the presentation of information already in trust will increase usage and facilitate the return of compensation to rights holders.

[0645] (c) Creating incentives for providing information For rights holders, entrusting their information creates an incentive to receive compensation, leading to the integration of more useful information into this information processing method.

[0646] (6-2) Differentiation of presentation methods This information processing method may differentiate between trusted information and non-trusted information on the observation space 220 using the following display methods (see Figure 6).

[0647] (a) Display of Trusted Badge For information that has been placed in trust, badges such as "Placed in Trust" and "Details Available" will be displayed.

[0648] (b) Display of level of detail For information that has been placed in trust, the level of detail is indicated as "Full text available," while for information that has not been placed in trust, it is indicated as "Summary only," etc.

[0649] (c) Differentiation of color tones Trusted information will be differentiated by color, such as green for trusted information and gray for non-trusted information.

[0650] (d) Adjusting transparency Similar to the visualization of confidence level 601 in Example 5, the transparency of the trusted information is set high (opaque), and the transparency of the non-trusted information is set low (transparent) (see Figure 11).

[0651] This differentiation allows users to visually understand which information is available in more detail.

[0652] (7) Reporting to rights holders and ensuring transparency

[0653] (7-1) Usage Report The information processing method according to this embodiment may include a step of reporting to the rights holder the actual usage status of their copyrighted works, etc.

[0654] For example, the report may include the following: trends in usage (trends in the number of views, citations, etc., along with time series 320, displayed using the time series transition observation function in Example 2), distribution of user attributes (distribution of user attributes such as industry, region, and company size, anonymized aggregated information), analysis of usage context (in what evaluation context it was used), aggregated evaluations (aggregated user evaluations), and basis for calculating compensation (details of the basis for calculating compensation based on actual usage) (see Figure 6).

[0655] The aforementioned reports may be provided in real time or periodically through a dashboard exclusively for rights holders.

[0656] (7-2) Ensuring Transparency The information processing method according to this embodiment may provide the following functions in order to ensure transparency in the calculation and distribution of consideration.

[0657] (a) Disclosure of calculation logic The method for calculating the consideration, the basis for setting the unit price, etc., will be disclosed to the rights holder.

[0658] (b) Verifiability of actual usage Rights holders will be able to verify the actual usage of their copyrighted works through log records, etc. However, the privacy of individual users will be protected.

[0659] (c) Provision of auditing functions By applying the third-party audit function from Example 5, it becomes possible to verify the calculation of consideration by an auditing body designated by the rights holder (see Figure 11).

[0660] (d) Acceptance of objections A contact point will be established for complaints regarding the calculation of compensation, and a recalculation or explanation will be provided based on the complaint.

[0661] Ensuring the aforementioned transparency will gain the trust of rights holders and create an incentive for more rights holders to enter into rights trust agreements.

[0662] (8) Social effects and contributions to industrial development

[0663] (8-1) Valuing Failure Cases The rights management and compensation distribution mechanism in this embodiment has significant social effects, particularly in valuing failure cases.

[0664] Traditionally, failure stories have been less likely to be shared because disclosing them carries significant reputational risks for those involved, and no compensation is usually offered.

[0665] In this embodiment, the sharing of failure cases is promoted as follows.

[0666] (a) Provision of consideration Even in cases of failure, compensation is distributed to rights holders based on actual usage, creating an incentive for disclosure.

[0667] (b) Anonymization options At the request of the rights holder, it is possible to publish the work anonymously, while keeping the author's name confidential. This allows for reduced reputational risk while still receiving compensation.

[0668] (c) Providing opportunities to learn from failures By extracting 302 failure transition patterns, failure cases can be systematically reused and will prove valuable as a learning resource for society as a whole (see Figure 7).

[0669] By valuing the aforementioned failure cases, it becomes possible to prevent similar failures from being repeated and contribute to the development of industry in society.

[0670] (8-2) Generalization of knowledge The rights management mechanism in this embodiment also contributes to the generalization of knowledge.

[0671] Traditionally, knowledge possessed only by experts (such as know-how for success and lessons learned from failures) has been difficult for a wide range of users to utilize, even when published in the form of books, papers, etc., due to access barriers (price, difficulty of obtaining, etc.).

[0672] In this embodiment, knowledge generalization is achieved as follows.

[0673] (a) Low-cost access Users can access the necessary knowledge through this information processing method. Since the fee is calculated based on actual usage, costs are incurred only to the extent necessary.

[0674] (b) Distillation of knowledge In Example 4, the distillation know-how extracted from publicly available information is generalized by the calibration means 103 and made available to a wide range of users (see Figures 2 and 9).

[0675] (c) Integration of knowledge Knowledge from different sources is integrated into the same latent space 210, making comparison and referencing easier (see Figure 5).

[0676] The generalization of the aforementioned knowledge will lower barriers to entry in knowledge-intensive industries (IT, biotechnology, consulting, etc.), promote competition, and accelerate the overall development of the industry.

[0677] (8-3) Enhancing the creative motivation of rights holders Appropriate compensation and rebates will boost the creative motivation of rights holders, creating a virtuous cycle in which more useful knowledge is generated and made public.

[0678] The aforementioned contribution to industrial development is one of the important social significances of this invention.

[0679] (9) Preprocessing and harmonization

[0680] (9-1) Purpose of harmonization If the publicly available information differs from the representation format of the latent space 210, the information processing method includes a harmonization step to convert the information into a format that can be placed in the latent space 210 (see Figure 5).

[0681] The objectives of the aforementioned harmonization are as follows: to enable comparison of information of different forms and scales on a common latent space 210; to establish a correspondence with the asset element system (201-205) of the intangible asset 200 defined in Example 1; and to convert the information into a format that can be supplied to the calibration process in Example 4 (see Figures 3 and 9).

[0682] (9-2) Methods of harmonization The aforementioned harmonization may include one or more of the following processes in the state representation means 100 (see Figure 2).

[0683] (a) Factor extraction Factors related to the creation and deterioration of intangible assets were extracted from natural language literature and mapped to the asset elements (elements 1-5: symbols 201-205) in Example 1 (see Figure 3).

[0684] (b) Time series The temporal progression is extracted from the case description and converted into a representation on the time axis 320 in Example 2 (see Figures 6 and 7).

[0685] (c) Extraction of causal structure The causal relationships between the factors are extracted and represented as the combination state 206 in Example 1 (see Figure 3).

[0686] (d) Feature quantity Using a pre-trained model (e.g., LLM), the literature description is converted into features in the latent space 210. The information processing model mentioned in Example 2 may also be used for this purpose.

[0687] (e) Normalization of the scale This process converts information described using different scales and units into coordinates on a common reference space.

[0688] (f) Adjustment with the eigencoordinate system of the evaluation body The publicly available information is projected onto the coordinate system of the intrinsic value 413 of the evaluation entity defined in Example 4 (see Figures 5 and 9).

[0689] (9-3) Adding metadata In this embodiment, the following metadata may be attached to the harmonization result.

[0690] (a) Source identifier Source information such as the title of the document, author, publication year, and URL.

[0691] (b) At the time of update The time when the information was written or acquired. This is recorded as update history 208 (see Figure 3).

[0692] (c) Reliability An index 207 indicating the reliability of the information source, the integrity of the information, or the uncertainty in the harmonization process. The confidence score 601 may be calculated using a method similar to the uncertainty evaluation of the external observer model 104 in Example 4 (see Figures 2 and 11).

[0693] (d) Difficulty level of harmonization An indicator that shows whether or not there is missing information, the degree of ambiguity, etc.

[0694] (e) Rights trust information The rights trust information defined in (4-3) above.

[0695] The metadata may be reflected on the observation space 220 as visual attributes such as line thickness, transparency, hue, and brightness in the display process described later (see Figure 6).

[0696] (10) Placement into latent space

[0697] (10-1) Principles of Placement The publicly available information after the harmonization is placed in the latent space 210, as detailed in Example 3, by the state representation means 100 (see Figures 2 and 5).

[0698] The important points in this embodiment are as follows:

[0699] (a) Neutral arrangement The labels of success and failure do not affect the arrangement itself in the latent space 210. The information is arranged neutrally based on the extracted asset elements (201-205) and the combination state 206 (see Figure 3).

[0700] (b) Elimination of criterion dependence The arrangement in the latent space 210 does not presuppose any specific criteria (norms, targets, etc.). The evaluation is calculated in a subsequent calibration process as the distance to the reference module 410 (see Figure 9).

[0701] (c) Integration with the eigenspace of the evaluator Publicly available information is placed in the same latent space 210 as the intangible assets 200 directly owned by the user. This allows for comparison and cross-referencing of the two (see Figure 5).

[0702] (10-2) Implementation of the layout The aforementioned configuration may be implemented in any of the following forms: a state vector representation (coordinates in a multidimensional space), a graph structure (a network representing the relationships between asset elements), a state transition matrix (a probabilistic representation representing changes over time), or a combination thereof.

[0703] (11) Calibration process for evaluation criteria

[0704] (11-1) The role of calibration The calibration process in this embodiment is a direct application of the concept introduced in Embodiment 4 and is performed by the calibration means 103 (see Figures 2 and 9).

[0705] The publicly available intangible asset information is supplied to the following calibration process described in Example 4.

[0706] (a) Calibration of market value based on the attributes of the evaluator The possibility of combining publicly available case studies with the valuation entity's existing asset portfolio is evaluated, and the valuation entity's unique value of 413 is calculated (see Figure 9).

[0707] (b) Calibration of credit and risk assessments based on normative standards The distance between the behavioral patterns in publicly available cases and the standard module 410 is evaluated to calculate credibility or risk (see Figure 9).

[0708] (11-2) Public information as an external observation instrument function In this embodiment, the publicly available intangible asset information is treated as a group of historical documents that constitute the external observer function in Examples 2 and 4, and is input into the external observer model 104 (see Figure 2).

[0709] Specifically, the following processes will be performed.

[0710] (a) Calculation of similarity ratio The similarity ratio defined in Example 3 is used to evaluate the similarity between the user's current state and the published example.

[0711] (b) Construction of a state transition model The state transition model in Example 3 is reinforced by the future state generation means 102 using time-series information included in the publicly available examples (see Figure 2).

[0712] (c) Support for generating branching scenarios Refer to the decision-making branching points in the published examples and use them to generate the state transition branch 301 of the future state 300 in Example 3 (see Figure 7).

[0713] (11-3) Evaluation by standard module The publicly available information is supplied to the reference module 410 defined in Example 4, and the following evaluation is performed (see Figure 9).

[0714] (a) Calculation of distance from the reference point The distance between the state or behavioral patterns of 200 intangible assets in publicly available examples and standard module 410 (standards and normative documents such as the UN Charter, national laws and regulations, and industry guidelines) is calculated.

[0715] (b) Generation of evaluation results Based on the aforementioned distance, it is determined whether the published case is "consistent" or "deviant" with respect to the criteria. It is important to note that this determination is generated not as a label attached to the input information itself, but as a relative relationship with the criteria module 410.

[0716] (c) Reflection in the value specific to the evaluator Using the aforementioned evaluation results, we estimate the change in the evaluator's intrinsic value 413 when a user adopts a strategy similar to the publicly available example (see Figure 9).

[0717] (12) Use of information after correction

[0718] (12-1) Application to future prediction The calibrated public information may be used by the future state generation means 102 to generate the future state 300 in Example 3 (see Figures 2 and 7).

[0719] Specifically, the following processes can be considered: Referencing the state transition patterns in publicly available examples, the probability of the user's future states occurring is corrected, and branching points in publicly available examples are referenced to present the decision-making options the user may face as state transition branch 301 (see Figure 7).

[0720] (12-2) Comparison display This information processing method may also compare and display the user's current status and publicly available examples on the observation space 220 in Example 3 using the visualization means 101 (see Figures 2 and 6).

[0721] In the comparative display described above, the following visual distinctions may be used: distinguish between user-specific information and publicly available information by line type (solid / dashed), line thickness, and color tone; express the confidence level of publicly available information (metadata in (9-3) above) by transparency; express the elapsed time since the information was updated by brightness; and highlight trusted information by a trusted badge.

[0722] (12-3) Integration with explanatory functions and legally appropriate presentation This information processing method may be integrated with the explanation request function introduced in Example 3.

[0723] If a user requests an explanation for state transition branch 301, this information processing method may present the publicly available examples referenced by the scenario and explain the similarity (similarity rate) (see Figure 7).

[0724] In this case, the following process is performed in accordance with the implementation of rights-compliant information presentation described in (3-4) above.

[0725] (a) Performing a conformity assessment For the publicly available cases presented, we will assess the risk of rights infringement.

[0726] (b) Selection of presentation method Based on the assessment results, one of the following methods will be selected: legal citation method, summary presentation method, structured information presentation method, link reference method, or full text presentation of entrusted information method.

[0727] (c) Attribution of source In accordance with Article 32 of the Copyright Act, the source of the quoted material is clearly indicated.

[0728] (d) Records of actual usage The usage record described in (5-1) above is executed as update history 208 and used as the basis for calculating the consideration (see Figure 3).

[0729] Through the aforementioned rights-compliant presentation, useful information can be provided to users without infringing on the rights of the rights holders.

[0730] (13) Reconstruction of historical conditions and counterfactual analysis

[0731] (13-1) Recreating the state at a past point in time In this embodiment, if the publicly available information describes the state at a specific point in the past, the information processing method may use that point in time as an initial condition and generate the future state 300 in Embodiment 3 using the future state generation means 102 (see Figures 2 and 7).

[0732] This allows for the following analysis.

[0733] (a) Historical verification We will compare the actual results from publicly available case studies with the prediction results obtained using this information processing method to verify the prediction accuracy.

[0734] (b) Counterfactual analysis We assume alternative options that were not adopted in the published examples and explore the possibility of different outcomes as state transition branches 301 (see Figure 7).

[0735] (13-2) Reuse as training data The results of the counterfactual analysis described above may be reused for the following purposes: to be stored as training data for the external observer model 104 in Example 4 (see Figure 2), to be used to improve the accuracy of the state transition model in Example 3, and to be used as the target for extracting distillation know-how in Example 4.

[0736] (14) User input of evaluations and updating of confidence levels

[0737] (14-1) Evaluation input function In this embodiment, the information processing method may also provide a function that allows the user to input an evaluation of publicly available information from the input unit 22 (see Figure 1).

[0738] The aforementioned evaluation input may include: expressions of agreement / doubt, comments, and evaluations of the user's contribution or usefulness to the results (star ratings, numerical ratings, etc.).

[0739] (14-2) Reliability update The aforementioned evaluation input may be used when updating the confidence level of the publicly available information (metadata in (9-3) above) (see Figure 11).

[0740] Specifically, the following processes are possible.

[0741] (a) Confidence update based on aggregation We will collect evaluations from multiple users and update the confidence score (601) of the publicly available information.

[0742] (b) Weighting based on predictability In Example 4, evaluations from users with a high predictability score are given higher weight.

[0743] (c) Calculation of provider's contribution We will compile evaluations of publicly available information provided by third-party providers and calculate the provider's contribution.

[0744] (d) Reflection in compensation Based on the evaluation-linked system described in (5-2) above, higher compensation will be calculated for the rights holders of information that receives a high evaluation.

[0745] (14-3) Visualization of evaluation The evaluation input and confidence score 601 may be visualized by the visualization means 101 on the observation space 220 detailed in Example 1 (see Figures 2 and 6).

[0746] For example, the following representations are possible: displaying the distribution of ratings as a heatmap; representing highly-rated public information as a stronger structure in the "fortress wall" metaphor of Example 2; and reducing the display priority of poorly-rated public information by increasing its transparency.

[0747] (15) Control of disclosed particle size

[0748] (15-1) Definition of Disclosure Levels In this embodiment, the information processing method may also provide a function to control the granularity of disclosure of publicly available information.

[0749] The aforementioned disclosure granularity may be defined at the following levels:

[0750] (a) Overview level Only summaries or key findings extracted from publicly available information will be displayed.

[0751] (b) level of distillation This section displays the generalized knowledge extracted as distillation know-how in Example 4.

[0752] (c) Level of detail (only information already in trust) This displays the original or detailed description of publicly available information. Only information based on a rights trust agreement is included.

[0753] (d) Factor level Only the factors extracted from publicly available information (asset elements 201-205 in Example 1) are displayed, and specific descriptions are hidden (see Figure 3).

[0754] (e) Anonymization level The source of the information is concealed, and only the 207 derived indicators are displayed.

[0755] (15-2) User selection The disclosed granularity may be dynamically selectable by the user from the input unit 22 (see Figure 1).

[0756] This offers the following benefits: Reduce cognitive load by displaying only an overview when detailed information is not needed; gain deeper insights by selecting the level of detail of the trusted information when detailed analysis is required; and utilize knowledge while protecting privacy by selecting the level of anonymization when the confidentiality of the source needs to be protected.

[0757] (15-3) Classification of information presentation formats based on whether or not compensation is involved The information processing method according to this embodiment may include a step of separating the form of information presentation depending on whether or not the presentation of publicly available intangible asset information to the user involves compensation to the rights holder.

[0758] Non-refundable forms include the presentation of statistical trends, generalized insights, or anonymized aggregate indicators 207 generated as a result of this information processing method learning and generalizing publicly available information.

[0759] Forms that involve the payment of consideration include quoting, summarizing, or detailing specific publicly available information, clearly indicating the source of such publicly available information, or providing the full text based on a rights trust agreement.

[0760] This separation ensures that the rights of rights holders are properly protected, and users can choose the level of detail of the information while being aware of whether or not there is a fee involved.

[0761] (16) Providing information to external systems

[0762] (16-1) API provided In this embodiment, calibrated public information or evaluation results based on such information may be provided to an external system via an API (Application Programming Interface) through the communication interface unit 13 (see Figure 1).

[0763] The API may provide the following information: a summary of publicly available examples for a specific intangible asset category, a list of publicly available examples similar to the user's current situation, the results of future scenarios based on the publicly available examples, and a flag for rights-cleared information (trusted information) (see Figure 7).

[0764] (16-2) Consideration of autoregressive effects As mentioned in Example 4, if an external entity uses the output of this information processing method to make a decision, that decision may affect the entire market and, consequently, the evaluation results of this information processing method.

[0765] In this embodiment, when considering the autoregressive effects, the information processing method may estimate the impact of providing information to an external system on the market and reflect this in the evaluation results. This reflection may also take into account importance level 502 (see Figure 10).

[0766] (17) State update via external trigger

[0767] In this embodiment, updates, additions, or deletions of publicly available intangible asset information are treated as external trigger information 500 and may be integrated into the state update mechanism triggered by an external trigger, as detailed in Embodiment 7 (see Figures 4 and 10).

[0768] Specifically, the following processes will be performed.

[0769] (a) Obtaining external trigger information The publication of new public information, the updating of existing information, or the deletion of information are obtained from the external information source 30 as external trigger information 500 (see Figures 1 and 10).

[0770] (b) Assessment of importance For the acquired external trigger information 500, the importance 502 to the intangible asset 200 or right to be evaluated is assessed (see Figure 10).

[0771] (c) Condition judgment If importance level 502 exceeds a predetermined threshold, this information processing method updates the coordinate values ​​of the relevant index 207 in the latent space 210.

[0772] (d) Generating and displaying differences The difference of 501 between the indicator 207 before and after the update is calculated and displayed on the observation space 220 by the visualization means 101 (see Figures 2 and 10).

[0773] This integration ensures that dynamic updates to publicly available information are reflected in real time, improving the overall evaluation accuracy of this information processing method.

[0774] (18) Clarification of the relationships between the examples This embodiment integrates and expands upon the concepts of Examples 1 to 5, as described below.

[0775] Relationship with Example 1: The asset element system (201-205) defined in Example 1 is used as the criteria for classifying and aligning publicly available information in this embodiment (see Figure 3). Furthermore, the concept of latent space 210 in Example 1 is extended in this embodiment to represent the space in which publicly available information is placed (see Figure 5).

[0776] Relationship with Example 2: The concept of standard documents introduced in Example 2 is extended in this embodiment to serve as a reference base for evaluating publicly available intangible asset information. Furthermore, the time-series transition observation function in Example 2 is used in this embodiment to reproduce past state changes in publicly available examples. The external observation function in Example 2 forms the basis for utilizing publicly available information as a historical document set in this embodiment.

[0777] Relationship with Example 3: The latent space 210 detailed in Example 3 functions in this embodiment as a space for uniformly arranging publicly available information (see Figure 5). Furthermore, the similarity rate and state transition model in Example 3 are applied in this embodiment to evaluate the similarity between publicly available examples and the user's current state and to predict future trends.

[0778] Relationship with Example 4: The calibration concept introduced in Example 4 forms the core of this embodiment (see Figure 9). Publicly available information is treated as a set of historical documents constituting the external observer function in Example 4 and is used for calculating the intrinsic value 413 of the evaluator and for evaluating the distance from the reference module 410. Furthermore, the regression learning of the decision path in Example 4 is applied in this embodiment to extract decision-making patterns in publicly available cases. The external observer model 104 in Example 4 forms the basis of the function that dynamically learns publicly available information in this embodiment (see Figure 2).

[0779] Relationship with Example 5: The concept of disclosure control detailed in Example 5 is applied in this embodiment to control the scope of use and display granularity of publicly available information. Furthermore, the audit function in Example 5 can be used in this embodiment as a means to verify the confidence level 601 of publicly available information, and as a means to ensure transparency in the calculation and distribution of consideration (see Figure 11).

[0780] Uniqueness of this embodiment: In this embodiment, we provide a technical means for handling not only intangible assets 200 directly owned by the user, but also knowledge contained in publicly available literature, case studies, or trained models, on a unified latent space 210 (see Figure 5).

[0781] Furthermore, this embodiment achieves the following by introducing a new technical configuration consisting of a rights management mechanism, a trust management mechanism, and a consideration calculation and distribution mechanism: lawful information presentation considering copyright law, unfair competition prevention law, publicity rights, portrait rights, etc.; creation of an incentive for information provision through appropriate compensation return to rights holders; promotion of effective sharing of not only success stories but also failure stories; and contribution to the democratization of knowledge and the development of industry.

[0782] This will enable even individual users with limited experience to evaluate and predict the future of intangible assets 200 by referencing the entire body of knowledge accumulated by humanity, and will also ensure that knowledge providers receive appropriate compensation, thereby forming a sustainable knowledge-sharing ecosystem.

[0783] (19) Effects As described above, the following effects can be obtained according to this embodiment.

[0784] (a) Generalization of knowledge By integrating publicly available intangible asset information and unifying it in the latent space 210, knowledge previously held only by specific experts becomes accessible to a wide range of users (see Figure 5).

[0785] (b) Consistency of evaluation By placing publicly available information and user-specific information on the same latent space 210 and evaluating them using a common criteria module 410, consistency of evaluation criteria is ensured (see Figure 9).

[0786] (c) Improvement of prediction accuracy By referring to the state transition patterns in the published examples, the accuracy of generating the future state 300 in Example 3 is improved (see Figure 7).

[0787] (d) Generalization of experience By generalizing the insights extracted from individual publicly available case studies as distillation know-how in Example 4, a knowledge base is formed that can be referenced by users facing similar situations.

[0788] (e) Making value from failure cases Even publicly available cases that resulted in failure are valued as knowledge about "under what conditions failure occurs" through distance evaluation against standard module 410, contributing to supporting users' decision-making (see Figure 9). Furthermore, the compensation distribution mechanism provides economic incentives to providers of failure cases, leading to the active sharing of lessons learned from failures that were previously difficult to share.

[0789] (f) Transparency and verifiability Evaluation results based on publicly available information can be verified by users if the source of the information is clearly indicated.

[0790] (g) Continuous improvement By reusing user-generated evaluations and counterfactual analysis results as training data for the external observer model 104, the accuracy of this information processing method can be continuously improved (see Figure 2).

[0791] (h) Visualization of reliability By managing the confidence level of publicly available information (601) as metadata and visually displaying it in the observation space (220), users can make decisions while considering the reliability of the information (see Figures 6 and 11).

[0792] (i) Protection of rights holders The rights of rights holders are appropriately protected by a rights management mechanism that takes into account copyright law, unfair competition prevention law, publicity rights, portrait rights, etc.

[0793] (j) Appropriate return of compensation A compensation calculation and distribution mechanism based on actual usage ensures that appropriate compensation is returned to rights holders, and that incentives for knowledge provision are sustainably maintained.

[0794] (k) Promoting information provision The incentive of reciprocal compensation leads to the integration of more useful information (both success and failure stories) into this information processing method, creating a virtuous cycle that improves evaluation and prediction accuracy.

[0795] (l) Contribution to industrial development The effective sharing of success and failure stories reduces the cost of trial and error for society as a whole, accelerates the pace of innovation, and promotes the development of knowledge-intensive industries.

[0796] (m) Integration with external triggers By treating updates to publicly available information as external trigger information 500 and displaying differences 501 based on importance 502, users can grasp the impact of changes in publicly available information in real time (see Figure 10).

[0797] The information processing method according to this embodiment builds upon the technological foundation established in Examples 1 to 5, adding new technological value such as the integration of publicly available knowledge, rights management, trust management, consideration distribution, and external trigger linkage. This expands the knowledge base for the valuation of intangible assets 200 and rights, and contributes to the development of industry in society through the formation of a sustainable knowledge-sharing ecosystem (see Figures 1, 2, 3, 5, 9, and 10). [Examples]

[0798] (Examples of intent change detection, state updates triggered by external triggers, evaluation function correction based on user behavior, inter-stakeholder interaction, and collective state output)

[0799] The following describes another embodiment of the information processing method according to the present invention.

[0800] This embodiment is based on the configuration described in Example 1, which places information regarding intangible assets 200 and associated rights in a latent space 210 (see Figures 3 and 5), the observation and manipulation of state transitions along the time axis 320 described in Example 2 (see Figure 6), the generation of future states 300 and presentation of branching scenarios described in Example 3 (see Figure 7), the calibration based on evaluation subject attributes and normative criteria described in Example 4 (see Figure 9), the disclosure control of decision path attributes described in Example 5 (see Figure 11), and the integration of publicly available intangible asset information described in Example 6. It further includes a configuration that dynamically updates the state transition model in response to changes in user intent, fluctuations in the external environment, and inter-subject interactions in the market.

[0801] (1) Detection of changes in intent and analysis of focus shifts

[0802] (1-1) Estimation of Intentional Vectors The information processing method according to this embodiment may include a step in which the processing unit 11 estimates the user's intent vector from the user's operation history, gaze information, transitions of selected objects, changes in comparison objects, or changes in condition inputs presented by the user from the input unit 22 (see Figures 1 and 2).

[0803] The intent vector is associated with at least one of the asset elements defined in Example 1 (elements 1-5: symbols 201-205), the asset properties of the latent space 210, or the evaluation subject attributes in Example 4, and is expressed as the dimension the user is currently focusing on, or the change in the degree of focus (see Figures 3 and 5).

[0804] (1-2) Time series changes in focal position This information processing method may also include a step of recording the time-series changes of the intention vector in the update history 208 as transitions along the time axis 320 detailed in Example 2 (see Figures 3 and 6).

[0805] This time-series change is treated as an indicator 207 that indicates a shift in the user's interest from one asset element to another, or a change in the evaluation subject attributes (static attributes or decision path attributes) in Example 4.

[0806] (1-3) Reflection in the state transition model The estimated intention vector is input to the state transition model detailed in Example 3 by the future state generation means 102 and may influence the weighting or selection of branching conditions in the generation of the future state 300 (see Figures 2 and 7).

[0807] Specifically, state transition branches 301 related to asset elements or market conditions that are of high interest to users may be calculated as having a higher probability of occurrence and displayed on the display unit 21 as "lines having a thickness or width corresponding to the probability of occurrence" in Example 3 (see Figures 1 and 7).

[0808] As a result, this information processing method can present branching scenarios that immediately reflect changes in the user's focus, even if those changes occur rapidly.

[0809] (2) State update via external trigger

[0810] (2-1) Obtaining external trigger information The information processing method according to this embodiment may include a step of obtaining external trigger information 500, such as policy changes, legal amendments, news reports, official statistics, corporate financial statements, international affairs, technical announcements, infrastructure development plans, court judgments, or market transaction information, from an external information source 30 (see Figures 1 and 10).

[0811] The external trigger information 500 is treated as a type of external environmental information introduced in Example 2 and indicates the external conditions to be evaluated. However, it should be noted that the external trigger information 500 differs in nature from the publicly available intangible asset information (reference knowledge used for evaluation) in Example 6.

[0812] The acquired external trigger information 500 may be automatically obtained via the communication interface unit 13 from API integration, RPA (Robotic Process Automation), external databases, reports published by organizations, or user-specified information sources as mentioned in Example 2 (see Figure 1).

[0813] (2-2) Evaluation of the importance of trigger information This information processing method may include a step in which the processing unit 11 evaluates the importance 502 of the impact on the acquired external trigger information 500 on the intangible asset 200 or right to be evaluated (see Figures 2 and 10).

[0814] The aforementioned importance score 502 may be calculated based on the correspondence with past similar events using a method similar to the similarity rate defined in Example 3.

[0815] If the evaluation based on the impact and the corresponding severity 502 exceeds a predetermined threshold, this information processing method may push the external trigger information 500 to the display unit 21 as an "automatic notification when damage to intangible assets 200 is foreseeable due to an event unexpected by the user" as mentioned in Example 2 (see Figures 1 and 10).

[0816] (2-3) Coordinate updates in latent space The external trigger information 500 is treated as a variable that influences the coordinate values ​​of market elements, rights elements, or environmental elements in the latent space 210, as detailed in Example 3 (see Figures 5 and 10).

[0817] This information processing method may update the future state 300 by updating the arrangement of the indicators 207 in the latent space 210 using the state representation means 100 in accordance with the external trigger information 500, and by re-executing the state transition model in Embodiment 3 using the future state generation means 102 (see Figures 2, 5, and 7).

[0818] This configuration ensures that risks or opportunities arising from rapid changes in the external environment are immediately reflected for users.

[0819] (3) Correction of evaluation function based on user behavior

[0820] (3-1) Acquisition of user behavior The information processing method according to this embodiment may include a step of acquiring the user's behavior history from the input unit 22 when the user selects, compares, adopts, saves, or rejects a specific future state 300 from among the presented state transition branches 301 (see Figures 1 and 7).

[0821] The aforementioned behavioral history is treated as part of the decision-making path attributes defined in Example 4 and may be subject to disclosure control in Example 5 (see Figure 11).

[0822] (3-2) Construction of the evaluation function The evaluation function in this information processing method may be configured as a weighting of at least one of the following: the importance of the asset elements (elements 1 to 5: symbols 201 to 205) defined in Example 1, the credibility evaluation based on the distance to the reference module 410 in Example 4, or the degree of responsiveness to the branching conditions of the state transition in Example 3 (see Figures 3 and 9).

[0823] (3-3) Correction based on behavioral history This information processing method may include a step of correcting the weighting of the evaluation function using the calibration means 103 based on the acquired user behavior history (see Figure 2).

[0824] Specifically, if a user repeatedly adopts the same type of scenario (for example, a scenario that emphasizes a particular asset element), the weighting coefficient corresponding to that asset element may be increased.

[0825] This correction may be implemented in the calculation of the subject-specific value 413 in Example 4 as a process that reflects the judgment path attributes (see Figure 9).

[0826] (3-4) Individualized branching presentation By using the corrected evaluation function, the state transition model is re-executed by the future state generation means 102, and in the subsequent generation of future states 300, branching conditions that are closer to the user's values ​​may be preferentially presented to the display unit 21 (see Figures 1, 2, and 7).

[0827] This allows users to more efficiently identify state transition branches 301 that align with their strategic intentions.

[0828] (4) Reflection of mutual influence between entities

[0829] (4-1) Definition of inter-subject dependency The information processing method according to this embodiment may include a step of representing the relationships in which multiple entities in the market (companies, organizations, nations, investor groups, or a collection thereof) can influence each other as dependencies between coordinates in the latent space 210 (see Figure 5).

[0830] The aforementioned dependency may include the following elements:

[0831] (a) Impact based on subject attributes and the severity level 502 The impact and importance of the subject based on attributes such as the history of exercise of rights, bargaining power, investment scale, market share, or element 5 (brand value: symbol 205) in Example 1 (see Figures 3 and 10).

[0832] (b) Causal correlation based on past cases Causal correlation based on records of past inter-subject interactions included in the reference document sets (standard document sets and historical document sets) in Examples 2 and 6.

[0833] (c) Autoregressive effects on the overall market An interaction model that reflects overall market trends, including the "autoregressive influence when an external entity makes a decision using the output of this information processing method" mentioned in Example 4.

[0834] (4-2) Identification of subject response using an external observer model This information processing method may use the external observer model 104 detailed in Example 4 to identify the input / output characteristics of the third-party entity being observed and estimate how that entity will respond to specific external trigger information 500 (see Figures 2 and 10).

[0835] The external observation instrument model 104 is continuously updated by sequential search and collection in Example 4, which can improve the estimation accuracy.

[0836] (4-3) Integration into state transition models The inter-subject dependency relationships described above may be integrated into the state transition model in Example 3 and considered by the future state generation means 102 in the generation of the future state 300 (see Figures 2 and 7).

[0837] Specifically, changes in the behavior of large entities, exercise of rights by risk-generating entities, or changes in partnership relationships with collaborating entities may be reflected in the state transition branch 301 as an impact on the state of the user's intangible asset 200 (see Figures 3 and 7).

[0838] (4-4) Visualization of Interactions The mutual influence between the entities may be visualized in the observation space 220 by the visualization means 101 on the display unit 21 using the following visual representation (see Figures 1, 2, 5, and 6).

[0839] This includes arrows or lines connecting the influencing entity and the affected entity, the thickness, color, or brightness of lines indicating the magnitude of the influence and its evaluation based on the severity 502, the selection of colors indicating the direction of the influence (positive or negative influence), and the representation of the entry route in the "insect infestation" metaphor exemplified in Example 2.

[0840] (5) Generation of aggregate state and API output

[0841] (5-1) Definition of set state The information processing method according to this embodiment may include a step in the processing unit 11 to generate a set state that integrates state vectors of multiple entities or future states 300 (see Figure 2).

[0842] The aforementioned collective state may be composed of any of the following units: regional units (prefectures, economic zones, nations, etc.), industry units (manufacturing, IT industry, medical industry, etc.), technology unit units (AI technology, biotechnology, environmental technology, etc.), and policy zone units (customs unions, free trade zones, unified regulatory zones, etc.).

[0843] The aforementioned aggregate state may be generated in the latent space 210 by aggregating, averaging, or extracting the state vectors of individual entities as representative values ​​(see Figure 5).

[0844] (5-2) Generation of set predictions This information processing method may generate a set prediction (statistical estimation of the future states of multiple entities) by applying the state transition model in Example 3 to the set state using the future state generation means 102 (see Figures 2 and 7).

[0845] The aforementioned set prediction may be expressed in the following form: the average future state 300 for the entire set, the variance or standard deviation within the set, the future state 300 corresponding to a specific percentile within the set (e.g., the 25th percentile, median, 75th percentile), and the detection of outliers or anomalies within the set.

[0846] (5-3) API output function This information processing method may also include an API (Application Programming Interface) output function via a communication interface unit 13 for providing the set state or set prediction to an external information processing device 10 (see Figure 1).

[0847] The aforementioned API output function is implemented as a form of "providing information to an external system" as mentioned in Example 6, and may provide the following information: the current value of the collective state in a specific region, industry, or technology field; future market trends based on collective forecasts; the relative position of the user within the collective (rank, standard score, etc.); and the time-series trend of the collective state (represented by a time axis of 320 in Example 2) (see Figure 6).

[0848] (5-4) Integration with external systems Through the aforementioned API output function, this information processing method may be linked with the following external systems: dashboard systems (visualization tools for managers), risk analysis systems (financial institutions, insurance companies, etc.), decision support tools (investment decisions, M&A evaluations, etc.), external evaluation systems (rating agencies, audit firms, etc.), and academic research platforms (for research purposes in economics, management, etc.).

[0849] This expands the scope of application of this information processing method from supporting individual users' decision-making to supporting market-wide analysis and policy formulation.

[0850] (6) Display and Operation Modes

[0851] (6-1) Visualization of intent vectors In this embodiment, the information processing method may visualize the intention vector estimated in (1) above using the visualization means 101 in the observation space 220 (see Figures 2, 5, and 6).

[0852] Specifically, the following display methods are possible: a cursor or highlighting to indicate the user's current point of focus; highlighting highly noteworthy asset elements using "shading or lighting" as in Example 1; and dynamically representing the time-series change of the focal point as "display updates along the time axis 320" as in Example 2 (see Figure 6).

[0853] (6-2) The impact of external triggers and the display of their importance. This information processing method may display the influence of the external trigger information 500 acquired in (2) above and the evaluation based on its importance 502 on the display unit 21 using the visualization means 101 in the following visual representation (see Figures 1, 2, and 10).

[0854] The magnitude of the impact and its evaluation by importance level 502 are represented as "number of insects, amount of movement, and intensity" in Example 2, the time of impact occurrence is placed as a marker on the time axis 320 in Example 2, and the extent of the impact's spread is represented as ripples or diffusion in the "three-dimensionally recognizable image" in Example 1 (see Figures 6 and 10).

[0855] (6-3) Visualization of Inter-stakeholder Influence The mutual influence between the subjects in (4) above may be visualized in the observation space 220 by the visualization means 101 in the following display manner (see Figures 2, 5, and 6).

[0856] The network diagram represents the main entities as nodes and the influence relationships as edges. The intensity of the influence (the magnitude of the influence and its evaluation by the importance scale 502) is represented by the thickness and color of the edges, or by the "lighting intensity" in Example 1. The temporal change in the intensity of the influence is represented by video techniques such as "fade in, fade out" in Example 2.

[0857] (6-4) Overview view of the grouped state The aggregated state in (5) above may be presented on the display unit 21 as an overview display similar to the "floating map window 310" in Embodiment 3 (see Figures 1 and 6).

[0858] This allows users to observe set predictions while understanding where their own state corresponds to within the overall set.

[0859] (6-5) Display in AR / VR environments The information processing method according to this embodiment may be executed by the user terminal 20 in the AR / VR environment mentioned in Embodiment 3 (see Figure 1).

[0860] In an AR / VR environment, the following operations become possible: estimation of intent vectors linked to eye movement, selection of 500 external trigger information via hand gestures, detailed display of inter-subject influence via proximity operations, and three-dimensional arrangement of group states within space.

[0861] This enables a display mode in which the importance level 502 of the state transition branch 301 changes in conjunction with the user's focus shift.

[0862] (7) Relationship between the examples

[0863] This embodiment is constructed based on the concepts of Embodiments 1 to 6, as described below.

[0864] Relationship with Example 1: The asset elements (elements 1-5: symbols 201-205) defined in Example 1 are used as corresponding dimensions for the intention vector in this embodiment (see Figure 3). Furthermore, the concept of the latent space 210 constructed across Examples 1 to 3 forms the basis for coordinate updates that reflect the external trigger information 500 in this embodiment (see Figures 5 and 10).

[0865] Relationship with Example 2: The concept of external environmental information introduced in Example 2 forms the basis of the external trigger information 500 in this embodiment (see Figure 10). Furthermore, the time-series transition observation function in Example 2 is used in this embodiment as a means to visualize the time-series transition of the intention vector and the influence of external triggers (see Figure 6). The external observation function in Example 2 functions as the evaluation platform for inter-stakeholder influence in this embodiment.

[0866] Relationship with Example 3: The state transition model detailed in Example 3 is extended in this embodiment to comprehensively reflect intention changes, external triggers, and inter-agent influences (see Figure 7). Furthermore, the concept of similarity rate in Example 3 is applied in this embodiment to the influence of external trigger information 500 and the magnitude of its evaluation by importance 502 (see Figure 10). The projection 222 onto the observation space 220 in Example 3 is used in this embodiment as a means of visualizing complex interactions (see Figure 5).

[0867] Relationship with Example 4: The concept of decision path attributes introduced in Example 4 forms the basis for acquiring user behavior and correcting the evaluation function in this embodiment (see Figure 9). Furthermore, the external observer model 104 in Example 4 is directly used in this embodiment for estimating the response of a third party (see Figure 2). The concept of calibration in Example 4 forms the theoretical basis for the evaluation function correction process in this embodiment.

[0868] Relationship with Example 5: The concept of disclosure control detailed in Example 5 is applied in this embodiment as a means of protecting the privacy of user behavior history (see Figure 11). Furthermore, the audit record 600 in Example 5 can be used in this embodiment as a means of verifying the confidence level 601 of the aggregate state (see Figure 11).

[0869] Relationship with Example 6: The integration function for publicly available intangible asset information introduced in Example 6 forms the basis for referencing records of past interactions between entities in this embodiment. Furthermore, the concept of API output in Example 6 forms the basis for the collective state output function in this embodiment.

[0870] Uniqueness of this embodiment: This embodiment provides a technical means for integrating three dynamic factors: changes in user intent, fluctuations in the external environment, and mutual influences between stakeholders in the market. This enables decision-making support that reflects in real time the value fluctuations of intangible assets 200 in a complex and dynamic market environment that cannot be captured by static evaluation models.

[0871] (8) Effects

[0872] As described above, the following effects can be obtained according to this embodiment.

[0873] (a) Real-time adaptability This enables the immediate generation of future states 300 that reflect changes in user intent, improving the flexibility of strategic decision-making (see Figure 7).

[0874] (b) Responsiveness to changes in the external environment By detecting external trigger information 500 and autonomously updating the state transition model, the freshness of predictions is maintained and the accuracy of risk management is improved (see Figure 10).

[0875] (c) Individualized decision support Based on user behavior, the evaluation function is calibrated by the calibration means 103, and individualized state transition branches 301 are presented, thereby providing support that is consistent with the user's unique values ​​(see Figures 2 and 7).

[0876] (d) Visualization of market interactions Advanced scenario comparisons reflecting the mutual influence of market and policy actors become possible within the observation space 220, improving the quality of decision-making in complex market environments (see Figure 6).

[0877] (e) Expansion of scope The external API output of aggregate data expands the scope of this information processing method from supporting individual user decision-making to becoming a market-wide analysis tool and a policy-making support system.

[0878] (f) Continuous improvement of predictability By combining this with regression learning of the decision path in Example 4, the individualization accuracy of the evaluation function improves as user behavior history is accumulated as update history 208, and predictability is continuously improved (see Figure 3).

[0879] (g) Creation of social and economic value Providing aggregate data improves overall market transparency and mitigates information asymmetry, thereby contributing to improved social and economic efficiency.

[0880] The information processing method according to this embodiment dramatically improves the practicality in valuing intangible assets 200 and rights by adding a new technical value of dynamic adaptability to the technical foundation established in Examples 1 to 6 (see Figures 1, 2, 3, 5, 6, 7, 9, and 10). [Examples]

[0881] (Examples of construction of multi-layered evaluation models, self-verification of evaluation accuracy, enhancement of explainability of evaluation results, and improvement of evaluation accuracy through inter-system collaboration)

[0882] The following describes yet another embodiment of the information processing method according to the present invention.

[0883] This embodiment is based on the calculation method for intangible assets 200 and associated rights indicators 207 described in Example 1 (see Figure 3), the observation of time-series transitions described in Example 2 (see Figure 6), the generation of future states 300 and projection 222 from latent space 210 to observation space 220 described in Example 3 (see Figures 5 and 7), the calibration based on evaluation subject attributes and normative criteria described in Example 4 (see Figure 9), the disclosure control of decision path attributes described in Example 5 (see Figure 11), the integration of publicly available intangible asset information described in Example 6, and the detection of change in intent and reflection of inter-subject mutual influence described in Example 7. It is an example that additionally incorporates a multi-layered evaluation structure to verify the reliability 601 of the evaluation model itself, a mechanism to enhance the explainability of the evaluation results, and a configuration to realize cooperative evaluation among multiple information processing systems (see Figures 1, 2, and 11).

[0884] (1) Construction of a multi-layered evaluation model

[0885] (1-1) Hierarchical structure of the evaluation layer The information processing method according to this embodiment may include a step of hierarchically configuring multiple evaluation layers in the processing unit 11 with respect to the intangible asset 200 or right to be evaluated (see Figure 2).

[0886] The aforementioned evaluation layer may be defined as the following hierarchy:

[0887] (Layer 1: Basic evaluation layer) This layer calculates basic indicators 207 based on the asset elements (elements 1-5: symbols 201-205) defined in Example 1 (see Figure 3). In this layer, the value or state that each individual asset element possesses is evaluated by the state representation means 100.

[0888] (Layer 2: Coupling and evaluation layer) This layer calculates an index 207 based on the combination state 206 of asset elements defined in Example 1 (see Figure 3). This layer evaluates the synergy effects or synergistic risks that arise from the combination of multiple asset elements.

[0889] (Third layer: Market evaluation layer) This layer, introduced in Example 1 and detailed in Example 4, performs market value calibration using calibration means 103 (see Figures 2 and 9). This layer performs corrections according to the market environment in which the evaluation target is placed and the attributes of the evaluation entity.

[0890] (Layer 4: Normative evaluation layer) This layer performs credit and risk assessment based on the normative standards introduced in Example 4 (see Figure 9). In this layer, the conformity of the standards to be evaluated is calculated by distance evaluation with the standard module 410.

[0891] (Layer 5: Dynamic evaluation layer) This layer reflects the change in intent, external trigger information 500, and inter-stakeholder interactions introduced in Example 7 (see Figure 10). This layer updates evaluation values ​​in response to temporal changes and environmental fluctuations.

[0892] (1-2) Definition of interlayer dependencies This information processing method may include a step of explicitly defining the dependencies between the evaluation layers in the processing unit 11 (see Figure 2).

[0893] Specifically, a bidirectional dependency may be established in which the evaluation results of a lower layer (Layer 1) are referenced as preconditions in the evaluation of a higher layer (Layer 2 and beyond), and the evaluation results of the higher layer are fed back into correcting the evaluation method of the lower layer.

[0894] The aforementioned dependency may be represented in the latent space 210 as an inter-layer transformation matrix or an inter-layer coupling function (see Figure 5).

[0895] (1-3) Calculation of stratified confidence levels This information processing method may include a step in which the processing unit 11 calculates the confidence level 601 of the evaluation result in each evaluation layer (see Figures 2 and 11).

[0896] The confidence level 601 may be calculated based on the following factors.

[0897] (a) Completeness of input information This index (207) indicates the extent to which the information necessary for evaluating the relevant segment has been obtained. Factors such as the percentage of missing values, the freshness of the information, and the reliability of the information source (601) are taken into consideration (see Figure 3).

[0898] (b) Model compatibility This index 207 indicates how well the evaluation model used in this layer (such as the information processing model mentioned in Example 2, or the external observer model 104 in Example 4) fits the target of evaluation (see Figure 2). Past prediction accuracy, similarity of application examples, etc., are taken into consideration.

[0899] (c) Stability of evaluation This index 207 indicates how stable the evaluation results are when small fluctuations occur in the input information. It may be calculated using methods such as sensitivity analysis or perturbation analysis.

[0900] (d) Interlayer consistency This is an index 207 that shows the consistency between the evaluation results of lower-level and higher-level groups. It considers factors such as the presence or absence of contradictions and the consistency of causal relationships.

[0901] The confidence level 601 may be expressed as a numerical value in the range of 0 to 1, similar to the metadata of the publicly available intangible asset information in Example 6, or as a graded evaluation such as "high," "medium," or "low" (see Figure 11).

[0902] (1-4) Calculation of overall confidence This information processing method may include a step in which the processing unit 11 integrates the confidence levels 601 of each evaluation layer to calculate the overall confidence level of the evaluation results (see Figures 2 and 11).

[0903] The overall confidence level may be calculated by a weighted average of the confidence levels of each layer (601), the minimum value, or by propagation calculations that take into account inter-layer dependencies.

[0904] If the overall confidence score falls below a predetermined threshold, the information processing method may display a warning to the user on the display unit 21 and recommend obtaining additional information, changing the evaluation model, or postponing the decision (see Figure 1).

[0905] (2) Self-verification mechanism for evaluation accuracy

[0906] (2-1) Construction of a record of correspondence between prediction and actual results The information processing method according to this embodiment may include a step of recording the correspondence between evaluation results generated in the past (particularly the prediction of the future state 300 in Embodiment 3) and actual results observed after a period of time as an update history 208 in the storage unit 12 (see Figures 1 and 3).

[0907] The aforementioned record may include the following information:

[0908] (a) Information at the time of prediction These include the state of the subject being evaluated at time t0 when the prediction was generated, the evaluation model used, the applied calibration conditions, and the prediction confidence level of 601 (see Figures 9 and 11).

[0909] (b) Prediction content These include the predicted value at future time point t1, the predicted distribution, the state transition branch 301, and the probability of each occurring (see Figure 7).

[0910] (c) Performance information This includes the state of the subject being evaluated as actually observed at future time point t1, the observation method, and the confidence level of the observation (601).

[0911] (d) Discrepancy information This includes the difference between the predicted value and the actual value, the direction of the deviation (overestimation or underestimation), and the magnitude of the deviation. This deviation may be calculated using the same method as the value deviation (Δ) 411 in Example 4 (see Figure 9).

[0912] The aforementioned record is based on the time-series transition observation function in Example 2 and explicitly maintains the temporal relationship between the predicted time point and the actual observation time point along the time axis 320 (see Figure 6).

[0913] (2-2) Calculation of evaluation accuracy index This information processing method may include a step in which the processing unit 11 calculates an index 207 indicating the accuracy of the evaluation model based on the prediction-actual correspondence record (see Figure 2).

[0914] The accuracy index may be calculated as any one or a combination of the following:

[0915] (a) Mean Absolute Error (MAE) This is the average of the absolute differences between predicted and actual values.

[0916] (b) Mean Squared Error (MSE) or Root Mean Squared Error (RMSE) This is the average of the squared errors between the predicted and actual values, or its square root.

[0917] (c) Coefficient of determination (R²) Metric 207 indicates the extent to which a predictive model can explain fluctuations in actual values.

[0918] (d) Precision of branch prediction This is the percentage of state transition branches 301 in Example 3 that were correctly predicted (see Figure 7).

[0919] (e) degree of calibration of probabilistic predictions This is an index 207 that shows the consistency between the predicted probability of occurrence and the actual frequency of occurrence. For example, it evaluates whether a group of events predicted to have a "70% probability of occurrence" actually occurs with a frequency of 70%.

[0920] The accuracy index may be calculated separately for each condition, such as the type of object being evaluated (patent rights, trademark rights, know-how, etc.), market sector, region, and time range.

[0921] (2-3) Detection of systematic bias This information processing method may include a step in which the processing unit 11 detects a systematic bias inherent in the evaluation model by analyzing the prediction-actual correspondence record (see Figure 2).

[0922] The aforementioned systematic bias refers to the tendency of an evaluation model to consistently overestimate or underestimate under specific conditions.

[0923] Specifically, the following biases may be detected:

[0924] (a) attribute-dependent bias The accuracy of evaluations tends to vary systematically depending on the attributes of the evaluating entity (industry, size, region, etc.).

[0925] (b) Time-dependent bias The accuracy of the assessment tends to systematically decrease or fluctuate depending on the length of the forecast period.

[0926] (c) External environment-dependent bias The accuracy of the evaluation tends to fluctuate systematically under specific external environmental conditions (such as economic boom or recession, or increased or decreased regulations).

[0927] (d) Asset component-dependent bias For specific asset elements (elements 1-5 in Example 1: symbols 201-205), the evaluation accuracy tends to differ systematically (see Figure 3).

[0928] The detection of the bias may be performed by statistical testing, regression analysis, or anomaly detection using an information processing model (e.g., LLM) as described in Example 2.

[0929] (2-4) Automatic correction of the evaluation model This information processing method may include a step of automatically correcting the evaluation model using the calibration means 103 based on the detected systematic bias (see Figure 2).

[0930] The aforementioned correction may be carried out in one of the following forms:

[0931] (a) Parameter adjustment Adjust the internal parameters (weight coefficients, thresholds, etc.) of the evaluation model.

[0932] (b) Addition of calibration function Add a correction function that applies only under specific conditions.

[0933] (c) Model switching Under specific conditions, switch to an alternative evaluation model.

[0934] (d) Changes to the ensemble configuration Change the weighting when combining multiple evaluation models.

[0935] It should be noted that the aforementioned correction is consistent with the calibration concept in Example 4 and is implemented as an improvement in the accuracy of the evaluation method, rather than a change in the evaluation criteria (see Figure 9).

[0936] (2-5) Recording of correction history This information processing method may include a step of recording the content of the correction, the reason for the correction, and the effect of the correction as an audit log 209 in the storage unit 12 when the correction of the evaluation model is performed (see Figures 1 and 3).

[0937] The correction history may be stored with signature information (digital signature, hash value, etc.) for tamper detection, similar to the audit record 600 detailed in Example 5 (see Figure 11).

[0938] This allows for tracking the evolution of the evaluation model, ensuring the reproducibility and verifiability of the evaluation results.

[0939] (3) Strengthening the explainability of evaluation results

[0940] (3-1) Hierarchical structure of explanatory information The information processing method according to this embodiment may include a step in the processing unit 11 for hierarchically structuring explanatory information for the evaluation results (see Figure 2).

[0941] The hierarchy of the explanatory information may be defined as follows:

[0942] (Level 1: Conclusion Level) This is a summary of the evaluation results (e.g., "high value," "risk detected," etc.).

[0943] (Second level: Evidence level) The main basis for the evaluation results is ("Value improved by combining element 2 (code 202) and element 5 (code 205)," "Deviation from the reference module 410 detected," etc.) (see Figures 3 and 9).

[0944] (Third level: level of detail) This section contains detailed information about the calculation process, data used, and applied calibration conditions for each evaluation tier.

[0945] (Fourth level: Reference level) This includes publicly available intangible asset information in Example 6, past cases referenced by the external observer model 104 in Example 4, and reference information such as the similarity rate with those cases (see Figure 2).

[0946] The aforementioned hierarchical structure allows users to selectively obtain explanations with an appropriate level of detail according to their level of understanding or needs.

[0947] (3-2) Contribution Analysis This information processing method may include a step in which the processing unit 11 calculates the contribution of input elements or intermediate representations that directly act on the intangible asset 200 or right index 207 (or the evaluation result based on said index 207) to be evaluated (see Figures 2 and 3).

[0948] In this information processing method, the contribution level may be set for the purpose of improving the efficiency of computing resources. Specifically, "impact" is defined as a concept (in a broad sense) that indicates the possibility that input elements or external conditions may cause changes in the state quantities of the intangible asset 200 or rights being evaluated (indicator 207, state transition branch 301 of future state 300, or evaluation results based thereon), regardless of whether such changes are adopted as a display or update process for the user (see Figures 3 and 7).

[0949] "Importance 502" is set up to control which information from the aforementioned impacts is presented or processed at what granularity (display resolution 401, highlighting threshold, aggregation level, log storage granularity, etc.) based on local evaluation perspectives defined by the industry, organization, group, region, or personal beliefs of the user (see Figures 8 and 10).

[0950] Regarding "contribution," it can be defined as a measure that quantifies the magnitude of the effect (sensitivity, contribution allocation, counterfactual difference, etc.) of input elements or intermediate representations (e.g., output of the evaluation layer, features, distance component from normative criteria) that directly act on the intangible asset 200 or rights indicator 207 being evaluated, among the aforementioned influences (i.e., a measure concerning elements that are recognized to have a direct involvement in the evaluation results).

[0951] Thus, with respect to "impact," "importance level 502," and "contribution," this information processing method employs a technique that divides the evaluation into stages. This has the effect of enabling the effective allocation of computing resources, for example, for evaluation targets where a direct contribution is expected rather than a broader impact (cases where direct contribution should be emphasized in terms of resolution).

[0952] The aforementioned contribution may be calculated by one of the following methods.

[0953] (a) Sensitivity analysis This tool calculates how much the evaluation results change when specific input elements (asset elements, market conditions, normative standards, etc.) are varied.

[0954] (b) SHAP (SHapley Additive exPlanations) values We apply a method for calculating the contribution of features in a machine learning model.

[0955] (c) Counterfactual analysis The evaluation result is calculated assuming a specific element is absent, and the difference between this result and the actual evaluation result is used as the contribution.

[0956] (d) Stratified contribution The contribution of each evaluation layer defined in (1) above is calculated.

[0957] The aforementioned contribution may be visualized on the observation space 220 detailed in Example 1 by the visualization means 101 as color intensity, line thickness, brightness, or by the highlighting function in Example 4 (see Figures 2, 5, and 6).

[0958] (3-3) Generating comparative explanations This information processing method may include a step in the processing unit 11 to generate comparative explanations in order to facilitate understanding of the evaluation results (see Figure 2).

[0959] The aforementioned comparative explanation may include the following forms:

[0960] (a) Contrast with counterfactuals This is an explanation such as, "If element 2 (symbol 202) is not connected, the evaluation value will decrease to XX" (see Figure 3).

[0961] (b) Comparison with alternative conditions This includes explanations such as, "If the evaluating entity is Company B instead of Company A, the evaluation value will change to XX."

[0962] (c) Comparison with past cases This involves comparing similar cases extracted from publicly available intangible asset information in Example 6 with the current subject of evaluation, and providing explanations such as, "In past case X, the result was XX, but the current situation differs in the respect of YY."

[0963] (d) Comparison with the standard The distance to the reference module 410 in Example 4 is shown, along with explanations such as "There is a deviation in XX points compared to the case where the standard is met" (see Figure 9).

[0964] The comparative explanation may be integrated with the explanation request function in Example 3 and displayed on the display unit 21 in response to user operations (mouse over, long press, etc.) (see Figure 1).

[0965] (3-4) Evaluation of the reliability of the explanation This information processing method may include a step in which the processing unit 11 evaluates the confidence level 601 of the generated explanatory information itself (see Figures 2 and 11).

[0966] The confidence level 601 in the above explanation may be calculated based on the following factors:

[0967] (a) Reliability of the supporting data The confidence level of the data supporting the explanation (input information, reference examples, etc.) is 601. The confidence level of the metadata in Example 6, which is 601, is referenced.

[0968] (b) Certainty of causality This is an index207 that indicates the extent to which the causal relationship asserted in the explanation is supported statistically or logically.

[0969] (c) Completeness of explanation This is an index 207 that indicates the extent to which the factors influencing the evaluation results are covered in the explanation.

[0970] (d) Consistency of explanation This index 207 indicates whether multiple explanations (from different hierarchies and perspectives) generated for the same evaluation result are mutually consistent.

[0971] The confidence level 601 of the above explanation is displayed on the display unit 21 along with the explanatory information, allowing users to make decisions while considering the certainty of the explanation (see Figures 1 and 11).

[0972] (3-5) Explanation generation using natural language This information processing method may include a step in which the processing unit 11 generates the evaluation results and the explanatory information as natural language sentences using the information processing model (e.g., LLM) mentioned in Example 2 (see Figure 2).

[0973] The aforementioned natural language sentence may be generated to have the following characteristics:

[0974] (a) Adaptation to user attributes The level of detail in explanations and the frequency of use of technical terms will be adjusted according to the user's expertise, industry background, language, etc.

[0975] (b) Contextual consistency Referencing the intent vector in Example 7, explanations related to the perspective the user is currently focusing on will be presented preferentially.

[0976] (c) Step-by-step refinement It provides an interactive explanation that starts with an overview and gradually adds details according to the user's requests.

[0977] (d) Diverse forms of expression In addition to text, it generates a comprehensive explanation that is linked to visual representations in Example 1, time-series graphs in Example 2, and diagrams of state transition branching 301 in Example 3 (see Figures 6 and 7).

[0978] (4) Collaborative evaluation across multiple systems

[0979] (4-1) Integration configuration of the evaluation system The information processing method according to this embodiment may include a configuration in which multiple information processing devices 10 (hereinafter referred to as the evaluation system) cooperate to perform the evaluation (see Figure 1).

[0980] The evaluation system may be configured as any of the following:

[0981] (a) Specialized system This system specializes in the types of asset elements (elements 1-5: symbols 201-205) in Example 1, such as patent evaluation systems, trademark evaluation systems, and know-how evaluation systems (see Figure 3).

[0982] (b) Regional systems These are systems specific to a particular region or jurisdiction, such as evaluation systems based on Japanese law or evaluation systems based on U.S. law.

[0983] (c) System by evaluation body These include internal company evaluation systems, evaluation systems operated by external rating agencies, and evaluation systems operated by public institutions.

[0984] (d) Systems by evaluation method These include systems specialized in quantitative evaluation, systems specialized in qualitative evaluation, systems that use statistical methods, and systems that use AI technology.

[0985] The multiple evaluation systems may have a communication interface that exchanges information with each other via the communication interface unit 13, similar to the API output function of the aggregate state in Embodiment 7 (see Figure 1).

[0986] (4-2) Integration of evaluation results This information processing method may include a step of integrating evaluation results obtained from the multiple evaluation systems in the processing unit 11 (see Figure 2).

[0987] The aforementioned integration may be carried out by one of the following methods:

[0988] (a) Weighted average A weighted average of the evaluation results is calculated by assigning a confidence level of 601 or a weighting based on the level of expertise of each evaluation system (see Figure 11).

[0989] (b) Ensemble learning A meta-learning model calculates an integrated evaluation using multiple evaluation results as input.

[0990] (c) Bayesian estimation The evaluation results from each evaluation system are used as observed values, and the posterior distribution is updated using Bayes' theorem.

[0991] (d) Voting or majority vote The results from multiple evaluation systems ("high rating," "low rating," etc.) will be combined and determined by voting or majority rule.

[0992] (e) Adopt the minimum or maximum value. In risk assessment, the most conservative assessment is adopted, while in valuation, the most optimistic assessment is presented as a reference; selective integration is performed according to the purpose.

[0993] The selection of the aforementioned integration method may be specified by the user from the input unit 22, or the information processing method may be automatically selected according to the characteristics of the object to be evaluated (see Figure 1).

[0994] (4-3) Detection of discrepancies in evaluation results This information processing method may include a step in which the processing unit 11 detects any significant discrepancies in the evaluation results obtained from the multiple evaluation systems (see Figure 2).

[0995] The detection of the aforementioned discrepancy may also be based on the following index 207.

[0996] (a) Standard deviation This occurs when the standard deviation of multiple evaluation results exceeds a predetermined threshold.

[0997] (b) range This occurs when the difference between the maximum and minimum values ​​exceeds a predetermined threshold.

[0998] (c) Direction of evaluation This occurs when some systems give a "high rating," while others give a "low rating."

[0999] If the aforementioned discrepancy is detected, this information processing method may perform the following actions.

[1000] (i) Warning display The display unit 21 warns the user that there is uncertainty in the evaluation results (see Figure 1).

[1001] (ii) Analysis of the cause of the discrepancy We will compare the evaluation methods, input information, and calibration conditions used by each evaluation system to identify the cause of the discrepancies.

[1002] (iii) Request for additional information We recommend obtaining any additional information necessary to resolve the discrepancy.

[1003] (iv) Referral to an expert If automated integration is difficult, the judgment of a human expert will be required.

[1004] The detection of the discrepancy and the analysis of its cause may be implemented by treating each evaluation system as an external observation target by the external observation model 104, based on the external observation function in Example 2 (see Figure 2).

[1005] (4-4) Mutual learning between evaluation systems This information processing method may include a configuration in which the multiple evaluation systems learn from each other.

[1006] The aforementioned mutual learning may be carried out in the following form:

[1007] (a) Feedback on prediction accuracy Similar to the regression learning of the decision path in Example 4, the prediction accuracy of each evaluation system is evaluated over time, and the evaluation method of the system that showed high accuracy is referenced by the other systems (see Figure 9).

[1008] (b) Distillation learning This involves transferring knowledge from a highly accurate evaluation system (teacher model) to another evaluation system (student model).

[1009] (c) Federated Learning Each evaluation system aggregates the results it has learned locally and updates the global evaluation model. In this process, the concept of disclosure control from Example 5 is applied to protect the confidential information of each system while enabling learning (see Figure 11).

[1010] (d) Exchange of evaluation methods The evaluation methods (algorithms, calibration methods, etc.) that demonstrate high accuracy under specific conditions will be shared among systems.

[1011] Through the aforementioned mutual learning, an improvement in evaluation accuracy, which is difficult to achieve with individual evaluation systems, is realized for the system as a whole.

[1012] (4-5) Visualization of inter-system communication This information processing method may also visualize the status of cooperation between the multiple evaluation systems on the observation space 220 in Example 3 using the visualization means 101 (see Figures 2, 5, and 6).

[1013] Specifically, the following display methods are possible.

[1014] (a) Network diagram Each evaluation system is placed as a node, and the information exchange paths are displayed as edges. The thickness of the edges indicates the frequency or importance of information exchange (see Figure 10).

[1015] (b) Evaluation distribution map The evaluation results from multiple evaluation systems are placed on the latent space 210 to visualize the variability or convergence of the evaluations (see Figure 5).

[1016] (c) Time series change diagram The time-series transition observation function from Example 2 is applied to show how the evaluation results of each evaluation system changed and converged along the time axis 320 (see Figure 6).

[1017] (d) Trust map The evaluation accuracy of each evaluation system (the accuracy index in (2-2) above) is displayed as a heatmap on the display unit 21 (see Figures 1 and 11).

[1018] This allows users to visually understand how multiple evaluation systems collaborate to perform the evaluation.

[1019] (5) Validation function for evaluation results

[1020] (5-1) Comparison with external benchmarks The information processing method according to this embodiment may include a step in the processing unit 11 to compare the evaluation results calculated by this method with external benchmarks (such as indicators 207 published by public institutions, industry standards, or evaluations by experts) (see Figure 2).

[1021] Based on the above comparison, the following verification may be performed.

[1022] (a) Validity of absolute value We will verify whether the absolute magnitude of the evaluation value is consistent with external benchmarks.

[1023] (b) Validity of relative ranking This study verifies whether the relative ranking of multiple evaluation targets aligns with external benchmarks.

[1024] (c) Validity of time series changes We will verify whether the temporal changes in evaluation values ​​(trends of increase or decrease) are consistent ...

Claims

1. An information processing method that calculates an index representing the state of intangible assets and enables the exploration of the temporal changes and future states of said index, A process for calculating an index representing the state of an intangible asset based on multiple asset elements that constitute an intangible asset belonging to an individual, corporation, organization, or group, and the state of combination of those asset elements. A step of arranging the calculated indicators in a multidimensional latent space, With respect to the indicators placed in the latent space, the steps include updating the position of the indicators in accordance with the passage of time, updates to information regarding the asset elements, or changes in external conditions, and maintaining the update history. The process involves converting indicators corresponding to multiple asset elements selected based on user instructions from among the indicators arranged in the aforementioned latent space into a human-understandable observation space, and visually displaying them on said observation space. The process of making the indicators displayed on the observation space viewable by specifying the state change along the time axis while maintaining the dimensional structure in the latent space, The process involves using the current arrangement of indicators in the aforementioned latent space as an initial condition, generating indicators corresponding to future states based on a state transition model, and presenting them as future scenarios including multiple state transition branches. A step that enables at least a part of the process of generating the state transition model or the indicator corresponding to the future state to be executed using an externally provided computation model or analysis module, or enables execution by referring to the output obtained by the externally provided computation model or analysis module, Includes, The step of setting up the plurality of state transition branches using the state transition model includes a step of calculating the distance or similarity between the indicators arranged in the latent space, and generating a state transition branch when the distance or similarity satisfies predetermined conditions, A step of setting or estimating the measurement resolution and measurement range related to the calculation of the distance or similarity, With respect to the calculation results of distance or similarity corresponding to a difference less than the measurement resolution, the steps include: displaying them as identical, or assigning them the same rank or displaying them within a rank range; A step of adding an error band to the calculation result of the distance or similarity and displaying it on the observation space, A step of displaying a warning for the calculation result of the distance or similarity outside the measurement range, An information processing method characterized by further including the following.

2. In the information processing method described in claim 1, The calculation of the aforementioned distance or similarity is A step of setting a reference module which is composed of a reference vector or reference region generated by analyzing a set of documents that are referenced as normative standards, A step of arranging the behavioral data or rights exercise status of the entity to be evaluated as a behavioral vector in the aforementioned latent space, A step of calculating the distance between the action vector and the reference module, The process involves evaluating the time-series changes in the aforementioned distance, evaluating it as substantial compliance if the distance is stable over time, and evaluating it as a decline in credibility if the distance is tending to increase over time. Includes, The measurement resolution is applied to the calculation of the distance between the action vector and the reference module, distances less than the measurement resolution are treated as having no significant deviation, and when the distance expands over time and exceeds the measurement resolution, it is calculated as credit risk or norm deviation risk. The process further includes estimating future distances based on past time-series changes in the aforementioned distance, under the assumption that the trend of change in said distance will continue in the future. An information processing method characterized by the following.

3. In the information processing method described in claim 2, In calculating the aforementioned behavior vector, if the subject exhibits different behaviors depending on the attributes of the trading partner or counterparty, the step of detecting the difference in such behaviors as the variance or multimodality of the behavior vector in the latent space, The process involves evaluating the detected behavioral heterogeneity as a deviation from the reference module, thereby visualizing it as a potential credit risk on the observation space. An information processing method characterized by further including the following.

4. In the information processing method described in claim 2, A step of calculating the bias of an externally provided computational model or analysis module, information processing model, or the entity performing the evaluation, used in the evaluation of the intangible asset, as the distance between the evaluation characteristics related to the bias and the reference module, The process of displaying the calculated distance on the observation space to clearly show the user the bias of the information provider on which the evaluation result of the intangible asset is based, An information processing method characterized by further including the following.

Images