Information processing device, information processing method, and program

Abstract

An information processing device, an information processing method, and a program that can reduce the effort required for managing content. [Solution] The information processing device comprises a reception unit that receives content, a content processing unit that assigns information identifying a physical address to the content received by the reception unit, stores the content in a content pool, and virtually classifies the content stored in the content pool, and a display control unit that displays on a display unit the virtual classification structure of the content virtually classified by the content processing unit.

Description

[Technical Field]

[0001] The present invention relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] The search server uses the input keywords to retrieve documents related to the keywords from the stored documents as search results. The search server provides the search results to the user's terminal device, allowing the user to check the search results on the terminal device. There is known a technique for classifying content hits in a search into clusters according to the degree of similarity (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2005-078245 Summary of the Invention [Problem to be solved by the invention]

[0004] Content is managed according to predetermined management rules, but creating these rules requires a significant amount of man-hours. Furthermore, once determined, the management rules may become outdated over time. Furthermore, when storing content, the user must set up checks such as access rights, which is time-consuming.

[0005] An object of the present invention is to provide an information processing device, an information processing method, and a program that can reduce the effort required for managing content. [Means for solving the problem]

[0006] One aspect of the present invention is an information processing device comprising: a reception unit that receives content; a content processing unit that assigns information identifying a physical address to the content received by the reception unit, stores the content in a content pool, and virtually classifies the content stored in the content pool; and a display control unit that displays the results of the virtual classification of the content by the content processing unit on a display unit.

[0007] One aspect of the present invention is an information processing method executed by a computer, which receives content, assigns information identifying a physical address to the received content, stores the content with the assigned physical address in a content pool, virtually classifies the content stored in the content pool, and displays the results of the virtual content classification on a display unit.

[0008] One aspect of the present invention is a program that causes a computer to accept content, assign information identifying a physical address to the accepted content, store the content with the assigned physical address in a content pool, virtually classify the content stored in the content pool, and display the results of the virtual content classification on a display unit. [Effects of the Invention]

[0009] According to the present invention, the effort required for managing content can be reduced. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 illustrates an example of an information processing apparatus according to an embodiment of the present invention. [Figure 2] FIG. 10 is a diagram illustrating an example of user attribute information. [Figure 3] FIG. 2 is a diagram illustrating an example of metadata of content. [Figure 4] FIG. 10 is a diagram illustrating an example of creator attribute information. [Figure 5]FIG. 2 is a flowchart showing an example of the flow of operations of the information processing device of the present embodiment. [Figure 6] FIG. 2 is a flowchart showing an example of the flow of operations of the information processing device of the present embodiment. [Figure 7] FIG. 10 is a diagram illustrating an example of a classification structure setting screen. [Figure 8] FIG. 2 is a flowchart showing an example of the flow of operations of the information processing device of the present embodiment. [Figure 9] FIG. 1 is a diagram illustrating an example of an N-dimensional vector space. [Figure 10] FIG. 2 is a flowchart showing an example of the flow of operations of the information processing device of the present embodiment. [Figure 11] FIG. 2 is a flowchart showing an example of the flow of operations of the information processing device of the present embodiment. [Figure 12] FIG. 2 is a flowchart showing an example of the flow of operations of the information processing device of the present embodiment. [Figure 13] FIG. 2 is a flowchart showing an example of the flow of operations of the information processing device of the present embodiment. [Figure 14] FIG. 10 is a flowchart showing another example of the flow of operations of the information processing device of the present embodiment. [Figure 15] FIG. 10 is a diagram illustrating an example of a content classification screen. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, an information processing apparatus, an information processing method, and a program according to an embodiment will be described with reference to the drawings. The embodiment described below is merely an example, and the embodiment to which the present invention is applied is not limited to the following embodiment. In all the drawings for explaining the embodiments, the same reference numerals are used for components having the same functions, and repeated explanations will be omitted. Furthermore, in this application, "based on XX" means "based on at least XX," and includes cases where it is based on other elements in addition to XX. Furthermore, "based on XX" is not limited to cases where XX is used directly, but also includes cases where it is based on XX that has been calculated or processed. "XX" is any element (for example, any information).

[0012] (Embodiment) (Information processing device) An information processing device 100 of this embodiment creates management rules for content such as documents and manages the content based on the created management rules. Fig. 1 is a diagram showing an example of the information processing device 100 of this embodiment. The information processing device 100 is realized by a device such as a personal computer, a server, a smartphone, a tablet computer, or an industrial computer.

[0013] The information processing device 100 accepts content. The information processing device 100 assigns information identifying a physical address to the accepted content and stores the content with the assigned information identifying the physical address in a content pool. Examples of information identifying a physical address include a physical address and an ID equivalent to a physical address. The following description will continue with the case where a physical address is used as an example of information identifying a physical address. The information processing device 100 virtually classifies the content stored in the content pool and displays the results of the virtual content classification on a display unit. An example of the results of the virtual content classification is a classification tree that hierarchically represents the content. The information processing device 100 also searches for content stored in the content pool based on, for example, a user specification. The information processing device 100 displays the results of the content search on a display unit.

[0014] The information processing device 100 includes a user interface (UI) 10, an account registration / update processing unit 20, a reception unit 25, a content pool 30, a content processing unit 70, and a display control unit 66. The content processing unit 70 includes a virtual content pool construction processing unit 40, a hierarchical structure generation agent 50, a pseudo metadata search agent 62, a natural language search agent 64, and a virtual content pool 80.

[0015] The UI 10 includes an account registration / update UI 12, a content pool UI 14, a classification display UI 16, and a natural language search UI 18. One example of the UI 10 includes a display unit. The virtual content pool construction processing unit 40 constructs a virtual content pool 80. The virtual content pool construction processing unit 40 includes a content pool crawler 42, a metadata generation agent 44, and an LLM (Large Scale Language Model) 4.

[0016] The virtual content pool 80 includes a user attribute database (DB) 81, a metadata DB 82, a usage log DB 83, a vector space generation agent 84, an embedding processing unit 85, and a content vector space storage unit 86.

[0017] The account registration / update UI 12 is a UI for the user U to register and update an account. The content pool UI 14 is a UI that allows the user U to display the content contained in the content pool 30 and perform predetermined processing. The classification display UI 16 is a UI for displaying content by classification. For example, a classification display UI may be created for each organization or user, and the displayed content may be viewable based on the user U's specifications. The natural language search UI 18 is a UI that receives a search prompt from the user U and allows the user U to perform a search using the received search prompt.

[0018] The account registration / update processing unit 20 acquires an account registration request from the account registration / update UI 12. The account registration / update processing unit 20 registers user attribute information in the user attribute DB 81 based on the user identification information and the like included in the acquired account registration request.

[0019] FIG. 2 is a diagram showing an example of user attribute information. As shown in FIG. 2, the example of user attribute information includes an employee number as an example of user identification information, a name, affiliation data, information indicating a position, and a job type code. Examples of affiliation data include a business headquarters name, a business division name, a department name, a section name, a job description, and a department code. Returning to FIG. 1, the explanation will continue.

[0020] After the user attribute information is registered in the user attribute DB81, when the account registration / update processing unit 20 receives an account update request from the account registration / update UI12, it updates the user attribute information registered in the user attribute DB81 based on the user identification information, etc., included in the received account update request.

[0021] The receiving unit 25 receives content and stores the received content in the content pool 30. Here, the content may be created by the information processing device 100, or may be created outside the information processing device 100 and input to the information processing device 100. When content is created outside the information processing device 100 and input to the information processing device 100, the content may be transmitted to the information processing device 100 from a terminal device (not shown) connected to the information processing device 100 via a communication network. In this case, in the information processing device 100, a communication unit (not shown) receives the content transmitted from the terminal device, and the reception unit 25 acquires the content received by the communication unit. Here, the communication network includes the Internet, a wide area network (WAN), a local area network (LAN), a public line, a provider device, a dedicated line, a wireless base station, etc. The reception unit 25 may or may not store the received content in the content pool 30. Below, the explanation will be continued assuming, as an example, a case where the reception unit 25 stores the received content in the content pool 30. The content pool 30 stores one or more pieces of content in the information processing device 100. The one or more pieces of content can be displayed on the UI 10. The content pool 30 assigns a physical address to each of the one or more pieces of content stored therein. Each of the one or more pieces of content is displayed on the UI 10 while remaining stored in the content pool 30. The content pool 30 notifies the virtual content pool construction processing unit 40 of the physical address assigned to the stored content.

[0022] The virtual content pool construction processing unit 40 constructs a virtual content pool 80. The content pool crawler 42 receives physical addresses notified from the content pool 30. The content pool crawler 42 also checks the update status of content in the content pool 30 and newly added content, and collects (crawls) the physical addresses. For example, the content pool crawler 42 polls the update status of content in the content pool 30 and newly added content, and collects the physical addresses. The metadata generation agent 44 acquires a physical address from the content pool crawler 42 and identifies content corresponding to the physical address from the content pool 30 based on the acquired physical address. The metadata generation agent 44 generates metadata from the identified content. For example, the metadata generation agent 44 generates metadata from the content using a large-scale language model (LLM) 4. The metadata generation agent 44 associates the generated metadata with content identification information and stores it in the metadata DB 82.

[0023] Fig. 3 shows an example of content metadata. For each category, attribute items, descriptions, a method for automatically extracting metadata using an AI agent, the number of characters, the number of embedding dimensions, subspaces, chunk embedding / categories, etc. are associated. As shown in FIG. 3, an example of content metadata includes categories such as basic information, content information, permission and access information, and tracking and management information.

[0024] Examples of basic information include the title, creator, creation date and time, last updater, and last update date, all of which are subject to vectorization. Examples of content information include version, summary, keywords, category, theme, project status, purpose of use, content type, content format, and number of pages. Of these, summary, keywords, category, theme, project status, purpose of use, content type, and content format are targeted for vectorization. The authority and access information includes access authority, confidentiality level, approver, and approval date. Of these, access authority and confidentiality level are the targets of vectorization. The tracking and management information includes the status, storage location, related content, expiration date (lifecycle), and content ID. Of these, the status, expiration date (lifecycle), and content ID are the targets of vectorization. The metadata generation agent 44 may ask the user for the missing information and acquire it. For example, the metadata generation agent 44 outputs information for inquiring about the missing information to the display control unit 66. The display control unit 66 acquires the information for inquiring about the missing information from the metadata generation agent 44 and displays the information for inquiring about the missing information on the UI 10. The metadata generation agent 44 acquires information entered by the user in response to the information for inquiring about the missing information displayed on the UI 10. This allows the metadata generation agent 44 to complete the metadata. Returning to FIG. 1, the explanation will continue.

[0025] The user attribute DB 81 stores the user attribute information registered by the account registration / update processing unit 20 . The metadata DB 82 stores the metadata stored by the metadata generation agent 44 . The usage log DB 83 stores the usage log of the user U. An example of a usage log is a search history, which includes information indicating the date and time of the search, a content ID (identification information), search user attributes, and content metadata information. When a natural language search is performed, the search history further includes information indicating the search conditions.

[0026] The vector space generation agent 84 acquires user attribute information from the user attribute DB 81, acquires metadata from the metadata DB 82, and acquires usage logs from the usage log DB 83. Based on the acquired user attribute information, metadata, and usage logs, the vector space generation agent 84 generates an N-dimensional vector space (N is an integer greater than 1) as metadata representing the characteristics of the content. For example, N may be the number of attribute items to be vectorized among the meta information included in the content. The N dimensions include the creator (department, employee number, name, etc.) and summary / abstract. An example of an author has 10 dimensions, and an example of an outline / summary has 768 x 6 dimensions (for 3,000 characters), assuming 125 overlapping characters in 500-character chunks. The vector space generation agent 84 stores the generated N-dimensional vector space in the content vector space storage unit 86.

[0027] The content vector space storage unit 86 stores the N-dimensional vector space stored by the vector space generation agent 84 . The embedding processing unit 85 acquires user attribute information from the vector space generation agent 84, and embeds the content into an N-dimensional vector space stored in the content vector space storage unit 86 based on the acquired user attribute information. After the embedding processing unit 85 embeds the content in the N-dimensional vector space, the vector space generation agent 84 derives inter-content distances between the multiple pieces of content embedded in the N-dimensional vector space stored in the content vector space storage unit 86. Based on the derived inter-content distances, the vector space generation agent 84 generates k (k is an integer greater than 1) clusters from the multiple pieces of content. For example, the vector space generation agent 84 generates k clusters using the k-means algorithm based on the inter-content distances.

[0028] The hierarchical structure generation agent 50 analyzes the metadata for each piece of content included in the cluster generated by the vector space generation agent 84, using the LLM5, for the content stored in the content vector space storage unit 86. The hierarchical structure generation agent 50 recommends the optimal folder (lowest level) based on the results of the metadata analysis. The hierarchical structure generation agent 50 displays a classification display UI16 on the UI10 to recommend the optimal folder. For example, the hierarchical structure generation agent 50 outputs information indicating the classification to recommend the most suitable folder to the display control unit 66. The display control unit 66 acquires the information indicating the classification from the hierarchical structure generation agent 50 and displays the information indicating the classification on the UI 10. The hierarchical structure generation agent 50 acquires information input by the user for the information indicating the classification.

[0029] The natural language search agent 64 acquires user identification information and a search prompt from the natural language search UI 18 and outputs them to the pseudo metadata search agent 62 . The pseudo metadata search agent 62 acquires user identification information and a search prompt from the natural language search agent 64. Based on the acquired user identification information, the pseudo metadata search agent 62 acquires user attribute information from the user attribute DB 81. The pseudo metadata search agent 62 generates pseudo user attribute information using the LLM 6. When the pseudo user attribute information required for the search is available, the pseudo metadata search agent 62 stores the pseudo user attribute information and generates pseudo metadata based on the stored pseudo user attribute information.

[0030] Based on the generated pseudo metadata, the pseudo metadata search agent 62 arranges pseudo user attribute information in an N-dimensional vector space in which multiple contents stored in the content vector space storage unit 86 are embedded. The pseudo metadata search agent 62 lists a predetermined number of contents in descending order of distance from the arranged pseudo user attribute information and derives their likelihood.

[0031] The natural language search agent 64 acquires information indicating a predetermined number of contents and the likelihood of each of the predetermined number of contents from the pseudo metadata search agent 62. The natural language search agent 64 outputs the acquired information indicating the predetermined number of contents and the likelihood of each of the predetermined number of contents to the display control unit 66. The display control unit 66 acquires information indicating a predetermined number of contents and the likelihood of each of the predetermined number of contents from the natural language search agent 64, and displays it on the natural language search UI 18.

[0032] In the information processing device 100, the vector space generation agent 84 may derive the distance between contents further based on the weight vector. For example, the vector space generation agent 84 may generate a vector space by dividing all the attribute weights c i In this case, the vector space generation agent 84 analyzes the distribution state for each attribute and determines the weight c iThis allows us to define the distance between contents by averaging the variations between each attribute. i is assumed to follow a normal distribution (mean μ, standard deviation σ). In this case, the average variance for all attributes is expressed by equation (1), and the weight c for attribute i is i is expressed by equation (2). σ i is the standard deviation of attribute i.

[0033]

number

number

[0034] In the information processing device 100, the vector space generation agent 84 may optimize classification depending on the creator's situation. For example, the vector space generation agent 84 creates the creator's attributes as M-dimensional vector data (M is an integer greater than 1). Fig. 4 is a diagram showing an example of creator attribute information. As shown in Fig. 4, the example of creator attribute information includes an employee number as an example of creator identification information, a name, affiliation data, information indicating a position, and an occupation code. Examples of affiliation data include the business headquarters name, business division name, department name, section name, business content, department code, job title, and job type code. For example, the creator attribute information may be represented by an employee number, job content, department code, job title, and job type code.

[0035] The vector space generation agent 84 further adds the created M-dimensional vector data to generate an (M+N)-dimensional vector space as metadata expressing the characteristics of the content. The vector space generation agent 84 stores the generated (M+N)-dimensional vector space in the content vector space storage unit 86. Returning to FIG. 1, the explanation will continue.

[0036] The embedding processing unit 85 embeds the content into an (N+M)-dimensional vector space stored in the content vector space storage unit 86 . Here, the vector space generation agent 84 may create, as attributes of the creator, relationships between organizations (departments), hierarchical relationships within a department, etc. For example, the vector space generation agent 84 uses a graph neural network (GNN) to create relationships between organizations (departments), hierarchical relationships within a department, etc. By specifying the creator's attributes and cutting the N-dimensional space on the specified creator's attributes, it is possible to generate a subclassification of the classification. This method makes it possible to display the classification that is most suitable for an organization or individual.

[0037] In the information processing device 100, the vector space generation agent 84 may acquire management rules. The vector space generation agent 84 digitizes the acquired management rules and generates an R-dimensional vector space (R is an integer greater than 1). For example, the management rules include confidentiality, access rights, compliance requirements, life cycle settings, etc. The management rules may be set by a user U. The vector space generation agent 84 further adds the created R-dimensional vector data to generate an (R+M+N)-dimensional vector space as metadata expressing the characteristics of the content. The vector space generation agent 84 stores the generated (R+M+N)-dimensional vector space in the content vector space storage unit 86.

[0038] The embedding processing unit 85 embeds the content into an (R+N+M)-dimensional vector space stored in the content vector space storage unit 86 . The embedding processing unit 85 acquires an image in which the content has been embedded in an (R+N+M)-dimensional vector space, and outputs the image to the display control unit 66. The display control unit 66 acquires the image in which the content has been embedded in the (R+N+M)-dimensional vector space from the embedding processing unit 85, and displays it on the UI 10. This makes it possible to visualize the distribution state in the (R+N+M)-dimensional, i.e., (R, N, M) vector space.

[0039] In the information processing device 100, the vector space generation agent 84 may optimize classification according to a time attribute. For example, the vector space generation agent 84 creates a time attribute as T-dimensional vector data (T is an integer greater than 1). An example of the time attribute may be a two-dimensional attribute consisting of an attribute for specifying the creation date or expiration date and a time indicating the corresponding date and time, i.e., (attribute, time). The vector space generation agent 84 further adds the created T-dimensional vector data to generate a (T+R+M+N)-dimensional vector space as metadata expressing the characteristics of the content. The vector space generation agent 84 stores the generated (T+R+M+N)-dimensional vector space in the content vector space storage unit 86. The embedding processing unit 85 embeds the content into a (T+R+N+M)-dimensional vector space stored in the content vector space storage unit 86 .

[0040] In the information processing device 100, the vector space generation agent 84 may create user attributes as U-dimensional vector data (U is an integer greater than 1). Since FIG. 4 can be applied to the user attributes, a description thereof will be omitted. The vector space generation agent 84 further adds the created U-dimensional vector data to generate a (U+T+R+M+N)-dimensional vector space as metadata expressing the characteristics of the content. The vector space generation agent 84 stores the generated (U+T+R+M+N)-dimensional vector space in the content vector space storage unit 86.

[0041] The embedding processing unit 85 embeds the content that the user U has searched for in the (U+T+R+N+M)-dimensional vector space stored in the content vector space storage unit 86. The vector space generation agent 84 calculates the distance between the content and specific content included in the search history. The pseudo metadata search agent 62 obtains search results based on the distance calculated by the vector space generation agent 84. For example, the pseudo metadata search agent 62 may preferentially obtain search results based on the shortest distance.

[0042] Here, the vector space generation agent 84 may calculate the distance between all the contents included in the search history or the contents included in the same category, instead of being limited to a specific content included in the search history. The pseudo metadata search agent 62 may have a first trained model that derives content features based on a combination of one or more pieces of content included in the content search results. The pseudo metadata search agent 62 creates information for presenting the derived content features and outputs it to the display control unit 66. The display control unit 66 causes the UI 10 to display the information for presenting the content features from the pseudo metadata search agent 62.

[0043] The first trained model is a model that has been machine-learned to understand the relationship between one or more content combinations included in content search results and content features derived based on the one or more content combinations. For example, the first trained model is created by machine learning using one or more content combinations included in content search results as explanatory variables and content features derived based on the one or more content combinations as objective variables.

[0044] Furthermore, the pseudo metadata search agent 62 may have a second trained model that derives a question for narrowing down the content based on the derived content features. The pseudo metadata search agent 62 creates a question for narrowing down the derived content and outputs it to the display control unit 66. The display control unit 66 causes the UI 10 to display the question for narrowing down the content from the pseudo metadata search agent 62. The second trained model is a machine-learned model of the relationship between the features of the content and the questions for narrowing down the content derived based on the features of the content. For example, the second trained model is created by machine learning using the features of the content as explanatory variables and the questions for narrowing down the content derived based on the features of the content as objective variables. This allows the creation of information to present the user with the characteristics of the content to be presented based on the content search results, and allows the user to ask questions to narrow down the search to the most suitable content. Based on the answers, the system can present more suitable content.

[0045] In the information processing device 100, the embedding processing unit 85 may embed content in a (T+R+N+M)-dimensional vector space stored in the content vector space storage unit 86, and present nearby content as similar documents to the user. Furthermore, the vector space generation agent 84 may present the reciprocal of the distance between contents to the user as the degree of similarity. In this case, if content with an inter-content distance of zero exists, the vector space generation agent 84 creates information for notifying the user U that there is a high possibility that the document is a duplicate, and outputs the information to the display control unit 66. The display control unit 66 may display information indicating that there is a high possibility that the document is a duplicate from the vector space generation agent 84 on the UI 10.

[0046] The vector space generation agent 84 may generate, as metadata expressing the characteristics of the content, a (R+N)-dimensional vector space, a (T+N)-dimensional vector space, a (U+N)-dimensional vector space, a (T+M+N)-dimensional vector space, or a (U+M+N)-dimensional vector space. In other words, the vector space generation agent 84 can generate a vector space of dimension N added to any combination of T, R, M, and U.

[0047] In the above-described embodiment, the functions of the information processing device 100 may be distributed and realized by a plurality of devices. In this case, the plurality of devices that distribute and realize the functions of the information processing device 100 may be directly connected to each other to input and output information, or may be connected via a communication network to transmit and receive information. Furthermore, among the plurality of devices that distribute and realize the functions of the information processing device 100, some devices are directly connected and some devices are connected via a network. All or part of the account registration / update processing unit 20, reception unit 25, content pool crawler 42, metadata generation agent 44, hierarchical structure generation agent 50, pseudo-metadata search agent 62, natural language search agent 64, display control unit 66, vector space generation agent 84 and embedding processing unit 85 are functional units (hereinafter referred to as software functional units) realized by a processor such as a CPU (Central Processing Unit) executing a program stored in a memory unit (not shown).

[0048] All or part of these functional units may be realized by hardware such as an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), or an FPGA (Field-Programmable Gate Array), or may be realized by a combination of software functional units and hardware.

[0049] The operation of the information processing device 100 will be described below with reference to FIGS. 5 is a flow diagram showing an example of the flow of operations of the information processing device 100 of this embodiment. The process of storing content by the information processing device 100 will be described with reference to FIGS. (Step S1-1) The information processing device 100 performs a login process. For example, the user U inputs user identification information and a password on an initial screen (not shown) displayed on the UI 10. The information processing device 100 performs the login process based on the input user identification information and password.

[0050] (Step S2-1) The information processing device 100 creates content. For example, a user U creates content using the information processing device 100. Here, the content may be input to the information processing device 100 from an external device. For example, the content may be transmitted to the information processing device 100 from a terminal device (not shown) connected to the information processing device 100 via a communication network. In this case, the information processing device 100 receives and acquires the content transmitted by the terminal device. (Step S3-1) The content pool UI 14 detects that the "Save As" button (not shown) has been pressed.

[0051] (Step S4-1) The receiving unit 25 receives the content and stores the received content in the content pool 30 . (Step S5-1) The content pool 30 notifies the content pool crawler 42 of the physical addresses assigned to the stored content. (Step S6-1) The content pool crawler 42 receives the physical address notified by the content pool 30 .

[0052] (Step S7-1) The metadata generation agent 44 acquires a physical address from the content pool crawler 42. Based on the acquired physical address, the metadata generation agent 44 identifies content corresponding to the physical address from the content stored in the content pool 30. The metadata generation agent 44 generates metadata from the identified content. The metadata generation agent 44 associates the generated metadata with content identification information and stores it in the metadata DB 82.

[0053] (Step S8-1) The hierarchical structure generation agent 50 analyzes the metadata for each piece of content included in the cluster generated by the vector space generation agent 84 for the content stored in the content vector space storage unit 86. The hierarchical structure generation agent 50 recommends an optimal folder (lowest level) based on the results of the metadata analysis. The hierarchical structure generation agent 50 displays a classification display UI 16 on the UI 10 to recommend the optimal folder.

[0054] (Step S9-1) The hierarchical structure generation agent 50 determines whether the user U has performed an operation on the classification display UI 16 in response to the recommendation. (Step S10-1) If the hierarchical structure generation agent 50 determines that the user U has not performed an operation in response to the recommendation on the classification display UI 16, it causes the classification display UI 16 for recommending the next most suitable folder to be displayed on the UI 10. The process proceeds to step S9-1. (Step S11-1) When the hierarchical structure generation agent 50 determines that the user U has performed an operation in response to the recommendation on the classification display UI 16, it opens the recommended folder on the classification display UI 16 and displays the file name of the corresponding content.

[0055] 5, the information processing device 100 may perform the following process. For example, in step S6-1, the content pool crawler 42 may accept a physical address notified by any method other than the physical address notified by the content pool 30. Furthermore, in step S7-1, if there is any missing information in the metadata generated from the identified content, the metadata generation agent 44 may ask the user U to fill in the missing information. This allows the metadata generation agent 44 to complete the metadata.

[0056] 6 is a flow diagram showing an example of the flow of operations of the information processing device 100 of this embodiment. The process of generating a hierarchical structure of classification by the information processing device 100 will be described with reference to FIGS. (Step S1-2) The hierarchical structure generating agent 50 starts a classification setting program. An example of the classification setting program is included in the hierarchical structure generating agent 50. (Step S2-2) The hierarchical structure generation agent 50 displays all folders in the content pool 30 .

[0057] (Step S3-2) The hierarchical structure generation agent 50 detects that the user U has selected a folder to be classified and pressed the "Execute" button (not shown). (Step S4-2) The hierarchical structure generation agent 50 displays the "classification structure setting screen" based on the pressed "execute" button.

[0058] 7 is a diagram showing an example of a classification structure setting screen. As shown in FIG. 7, the example of the classification structure setting screen includes the total number of contents to be classified M, the maximum number of layers m, the average number of contents n1 contained in the folder at the bottom layer, the number of folders (number of classifications) k1 (=[M / n1]+1) at the bottom layer, and the number of folders n included in any i-th layer from the bottom layer. iand the number of folders in the i-th layer, k i can be set. Among these, the total number of contents to be classified is M, the number of folders in the lowest layer is n1, and the number of folders in the i-th layer is n i is set by the information processing device 100. The user U has the maximum number of layers m, the average number of contents n1 contained in the folders in the lowest layer, and the number of folders k contained in an arbitrary i-th layer from the lowest layer. i can be set by the user U.

[0059] If l major categories and m medium categories are set, folders for the medium categories are created as follows: The average number of folders contained in the folder one level above is n2. The average number of folders is entered on UI10. In this case, the number of categories is calculated using formula (4). k2=[k1 / n2]+1 (4) The final center of gravity (k) of the lowest folder obtained by the k-means method is divided into k2 folders using the k-means method to generate folders for intermediate classification. The user U sets the number of layers m in the hierarchical structure in advance, and the process ends when the number of layers reaches m.

[0060] User U refers to the classification structure setting screen and determines the number of folders n included in each hierarchy i. i The hierarchical structure generation agent 50 inputs the number of folders n included in each hierarchy i input by the user U (n1 for the lowest hierarchy) and the maximum hierarchy number m. i (In the case of the lowest layer, the number of contents n1) and the maximum number of layers m are set. Returning to FIG. 6, the explanation will continue.

[0061] (Step S5-2) The hierarchical structure generation agent 50 sets the hierarchy i to 1. (Step S6-2) The hierarchical structure generation agent 50 calculates the number of folders included in the ith folder from the bottom layer using the following formula (5). k i =[ki-1 / n i ]+1 (5) However, k0=M (total number of contents).

[0062] (Step S7-2) The hierarchical structure generating agent 50 determines whether or not the hierarchical level i is m. If the hierarchical structure generating agent 50 determines that the hierarchical level i is m, it determines that the hierarchical structure of the classification is complete. (Step S8-2) If the hierarchical structure generating agent 50 determines that each hierarchical level i is not m, it sets i to i+1 and proceeds to step S6-2.

[0063] In the information processing device 100, the hierarchical structure generation agent 50 may automatically improve the hierarchical structure. Specifically, the hierarchical structure generation agent 50 may restructure folders when the distance between the centers of gravity of folders in the lowest layer falls below an initially set threshold. Furthermore, the hierarchical structure generation agent 50 may restructure folders when the number of contents in the lowest layer folder exceeds the initially set value of n1 within an initially set range. The hierarchical structure generation agent 50 may restructure folders based on the number of views. By configuring in this way, folders can be restructured when the initially set hierarchical structure is no longer optimal as the amount of stored content increases or decreases.

[0064] Fig. 8 is a flow diagram showing an example of the flow of operations of the information processing device 100 of this embodiment. A process in which the information processing device 100 starts generating an optimal folder (lowest layer) will be described with reference to Figs. 1 and 8. Here, as an example, a case will be described in which the vector space generation agent 84 generates an N-dimensional vector space. (Step S1-3) The vector space generation agent 84 collects data on all files to be classified from the user attribute DB 81, the metadata DB 82, and the usage log DB 83.

[0065] (Step S2-3) The vector space generation agent 84 generates an N-dimensional vector space based on the acquired user attribute information, metadata, and usage log. The vector space generation agent 84 stores the generated N-dimensional vector space in the content vector space storage unit 86. The embedding processing unit 85 acquires content identification information and metadata from the vector space generation agent 84, and embeds the content into an N-dimensional vector space stored in the content vector space storage unit 86 based on the acquired content identification information and metadata.

[0066] (Step S3-3) The vector space generation agent 84 extracts k1 points from the N-dimensional space. For example, the vector space generation agent 84 may extract the k1 points from the N-dimensional space according to a predetermined rule, or may extract them randomly. (Step S4-3) The vector space generation agent 84 calculates the inter-content distance between each content and the extracted k1 points. For example, the vector space generation agent 84 calculates the inter-content distance according to the definition of the inter-content distance shown in Equation (6).

[0067]

number

[0068] The vector space generation agent 84 identifies the point with the shortest inter-content distance from among the points extracted in step S3-3, and collects the contents close to the identified point to generate k1 clusters.

[0069] (Step S5-3) The vector space generation agent 84 calculates a centroid point for each cluster, measured based on the inter-content distance. (Step S6-3) The vector space generation agent 84 calculates the distances between all the contents and the k1 centroids, and moves the contents that are close to the centroids of clusters different from the cluster to which they belong to the closer cluster.

[0070] (Step S7-3) The vector space generation agent 84 calculates the number of contents N transferred between clusters. T Determine whether or not N is equal to 0. T If N is the number of contents moved between clusters, the optimal folder generation is completed. T If it is not 0, the process moves to step S6-3.

[0071] Fig. 9 is a diagram showing an example of an N-dimensional vector space. As shown in Fig. 9, content is embedded in the example of the N-dimensional vector space. Fig. 9 also shows clusters C1 to C6 and centers of gravity COG1 to COG6 of clusters C1 to C6.

[0072] Fig. 10 is a flow diagram showing an example of the flow of operations of the information processing device 100 of this embodiment. A process in which the information processing device 100 starts generating an optimal folder (upper layer) will be described with reference to Fig. 1 and Fig. 10. Here, as an example, a case will be described in which the vector space generation agent 84 generates an N-dimensional vector space. (Step S1-4) The vector space generation agent 84 extracts k1 to k2 centroid points of the cluster immediately below. For example, the vector space generation agent 84 may extract k1 to k2 centroid points of the cluster immediately below according to a predetermined rule, or may extract them randomly. (Step S2-4) The vector space generation agent 84 measures the distance between the extracted k2 points and the centroid points of the lower clusters, and generates k2 clusters using the k-means method.

[0073] (Step S3-4) The vector space generation agent 84 calculates a centroid point for each cluster, measured based on the inter-content distance. (Step S4-4) The vector space generation agent 84 calculates the distances between all folders and the k2 centroids, and moves folders that are close to the centroids of clusters other than the one to which they belong to, to the nearest cluster.

[0074] (Step S5-4) The vector space generation agent 84 calculates the number of contents N that have been moved between clusters. T Determine whether or not N is equal to 0. T If N is the number of contents moved between clusters, then the generation of the optimal top folder is completed. T If it is not 0, the process proceeds to step S4-4.

[0075] Fig. 11 is a flow diagram showing an example of the flow of operation of the information processing device 100 of this embodiment. A process of natural language search performed by the information processing device 100 will be described with reference to Fig. 1 and Fig. 11. Here, as an example, a case will be described in which a user U is logged in and the vector space generation agent 84 is generating a (T+R+M+N)-dimensional vector space.

[0076] (Step S1-5) The natural language search UI 18 accepts a search prompt input by the user U. The natural language search agent 64 acquires the search prompt accepted by the natural language search UI 18 and the logged-in user identification information. The natural language search agent 64 acquires user attribute information from the user attribute DB 81 based on the acquired user identification information. The natural language search agent 64 creates a prompt creation request including the user identification information and the user attribute information, and outputs the request to the pseudo metadata search agent 62.

[0077] (Step S2-5) The pseudo-metadata search agent 62 receives a prompt creation request from the natural language search agent 64. The pseudo-metadata search agent 62 obtains the user identification information and user attribute information contained in the received prompt creation request, and generates a prompt using the LLM 6 based on the obtained user attribute information. The LLM 6 includes information that has been trained for pseudo-metadata generation. (Step S3-5) The pseudo metadata search agent 62 checks whether the user attribute information required for the search is available using the LLM 6. The LLM 6 includes a learning module that checks the availability of the user attribute information required for the search and lists the required user attribute information. For example, the pseudo metadata search agent 62 outputs information for inquiring about missing user attribute information to the display control unit 66. The display control unit 66 acquires the information for inquiring about missing user attribute information from the pseudo metadata search agent 62 and displays the information for inquiring about missing user attribute information on the UI 10. The pseudo metadata search agent 62 acquires information input by the user U in response to the information for inquiring about missing user attribute information displayed on the UI 10.

[0078] (Step S4-5) The pseudo metadata search agent 62 determines whether or not the user attribute information required for the search is available. (Step S5-5) If the pseudo metadata search agent 62 determines that the user attribute information required for the search is not available, it requests additional information from the user U. For example, the pseudo metadata search agent 62 may request the additional information via chat. Then, the process proceeds to step S1-5.

[0079] (Step S6-5) When the pseudo metadata search agent 62 determines that the user attribute information required for the search is available, it stores the pseudo user attribute information and generates pseudo metadata based on the stored pseudo user attribute information. (Step S7-5) Based on the generated pseudo metadata, the pseudo metadata search agent 62 places the pseudo metadata in a (T+R+M+N)-dimensional vector space in which multiple contents stored in the content vector space storage unit 86 are embedded.

[0080] (Step S8-5) The pseudo-metadata search agent 62 lists a predetermined number of contents in order of proximity to the placed pseudo-metadata and derives their likelihood. When the coordinates of a point in the content vector space generated by the pseudo-metadata are P, the likelihood L of content X is expressed, for example, by Equation (7). L(x)=α(1 / D(P,X)) (7) In equation (7), α is an arbitrary constant for adjusting the value. According to equation (7), it can be seen that the closer (smaller) the distance D is, the larger the likelihood L becomes.

[0081] The natural language search agent 64 acquires information indicating a predetermined number of contents and the likelihood of each of the predetermined number of contents from the pseudo metadata search agent 62. The natural language search agent 64 outputs the acquired information indicating the predetermined number of contents and the likelihood of each of the predetermined number of contents to the display control unit 66. The display control unit 66 acquires the information indicating the predetermined number of contents and the likelihood of each of the predetermined number of contents from the natural language search agent 64 and displays it on the UI 10.

[0082] Fig. 12 is a flow diagram showing an example of the operation flow of the information processing device 100 of this embodiment. Another process of natural language search performed by the information processing device 100 will be described with reference to Fig. 1 and Fig. 12. Here, as an example, a case will be described in which a user U is logged in and the vector space generation agent 84 is generating a (T+R+M+N)-dimensional vector space.

[0083] (Step S1-6) The natural language search UI 18 accepts a search prompt input by the user U. The natural language search agent 64 acquires the search prompt accepted by the natural language search UI 18 and the logged-in user identification information. The natural language search agent 64 acquires user attribute information from the user attribute DB 81 based on the acquired user identification information. The natural language search agent 64 creates a prompt creation request including the user identification information and the user attribute information, and outputs the request to the pseudo metadata search agent 62.

[0084] (Step S2-6) The pseudo-metadata search agent 62 receives a prompt creation request from the natural language search agent 64. The pseudo-metadata search agent 62 obtains the user identification information and user attribute information contained in the received prompt creation request, and generates a prompt using the LLM 6 based on the obtained user attribute information. The LLM 6 includes information that has been trained for pseudo-metadata generation.

[0085] (Step S3-6) The pseudo metadata search agent 62 checks whether the user attribute information required for the search is available using the LLM 6. The LLM 6 includes a learning module that checks the availability of the user attribute information required for the search and lists the required user attribute information. For example, the pseudo metadata search agent 62 outputs information for inquiring about missing user attribute information to the display control unit 66. The display control unit 66 acquires the information for inquiring about missing user attribute information from the pseudo metadata search agent 62 and displays the information for inquiring about missing user attribute information on the UI 10. The pseudo metadata search agent 62 acquires information input by the user U in response to the information for inquiring about missing user attribute information displayed on the UI 10.

[0086] (Step S4-6) If the generation of an attribute is not complete, the pseudo metadata search agent 62 excludes the dimension related to the attribute whose generation is not complete from the search space. (Step S5-6) The pseudo-metadata search agent 62 stores the user attribute information after excluding from the search space the dimensions related to attributes whose generation has not been completed in step S4-6, or when the generation of attributes has been completed in step S3-6, and generates pseudo-metadata based on the stored user attribute information.

[0087] (Step S6-6) Based on the generated pseudo metadata, the pseudo metadata search agent 62 places the pseudo metadata in a (T+R+M+N)-dimensional vector space in which multiple contents stored in the content vector space storage unit 86 are embedded. The pseudo metadata search agent 62 generates a subspace containing only the dimensions in which attribute data exists, and searches this space.

[0088] (Step S7-6) The pseudo metadata search agent 62 lists a predetermined number of contents in descending order of distance from the placed pseudo metadata and derives their likelihoods. The natural language search agent 64 acquires information indicating the predetermined number of contents and the likelihood of each of the predetermined number of contents from the pseudo metadata search agent 62. The natural language search agent 64 outputs the acquired information indicating the predetermined number of contents and the likelihood of each of the predetermined number of contents to the display control unit 66. The display control unit 66 acquires the information indicating the predetermined number of contents and the likelihood of each of the predetermined number of contents from the natural language search agent 64 and displays it on the UI 10.

[0089] Fig. 13 is a flow diagram showing an example of the flow of operation of the information processing device 100 of this embodiment. A process in which the information processing device 100 displays an optimal classification for the user U will be described with reference to Fig. 1 and Fig. 13. Here, as an example, a case will be described in which the vector space generation agent 84 generates a (T+R+M+N)-dimensional vector space. (Step S1-7) The information processing device 100 performs a login process. For example, the user U inputs user identification information and a password on an initial screen (not shown) displayed on the UI 10. The information processing device 100 performs the login process based on the input user identification information and password.

[0090] (Step S2-7) The display control unit 66 causes the UI 10 to display a category creation screen. The category creation screen includes a display for selecting whether to target content created by the user (user U), content created by others other than user U, or all content.

[0091] First, a case where the user U selects to target the content he or she has created will be described. (Step S3-7-1) The vector space generation agent 84 creates a subspace in a (T+R+M+N)-dimensional vector space in which multiple contents stored in the content vector space storage unit 86 are embedded, by fixing the creator attribute, which is M-dimensional vector data, to itself. (Step S4-7-1) The hierarchical structure generation agent 50 creates and completes a classification on this subspace.

[0092] Next, a case where the user U selects to target content created by others will be described. (Step S3-7-2) The vector space generation agent 84 creates a subspace in a (T+R+M+N)-dimensional vector space in which multiple contents stored in the content vector space storage unit 86 are embedded, by fixing the creator attribute, which is M-dimensional vector data, to not oneself (other than oneself).

[0093] (Step S4-7-2) The hierarchical structure generation agent 50 creates a classification on this subspace. (Step S5-7-2) The hierarchical structure generation agent 50 leaves the lowest folder containing previously viewed content based on the content viewing history as is, and creates a classification called "Other" for the other folders, moves them under this folder, and completes the process. By classifying the categories with access counts below a preset threshold as "Other," the hierarchical structure generation agent 50 can improve visibility.

[0094] Next, a case where the user U selects to target all content will be described. (Step S3-7-3) The vector space generation agent 84 creates and completes classification in a (T+R+M+N) dimensional vector space in which multiple contents stored in the content vector space storage unit 86 are embedded.

[0095] In the information processing device 100, the vector space generation agent 84 may use LLM to extract creator attributes that are similar to the user attributes of the user himself / herself. The vector space generation agent 84 may create a classification based on a subset of the content vector space based on the extracted creator attributes.

[0096] Fig. 14 is a flow diagram showing another example of the flow of operations of the information processing device 100 of this embodiment. A process in which the information processing device 100 displays an optimal classification for the user U will be described with reference to Fig. 1 and Fig. 14. Here, as an example, a case will be described in which the vector space generation agent 84 generates a (T+R+M+N)-dimensional vector space. (Step S1-8) The information processing device 100 performs a login process. For example, the user U inputs user identification information and a password on an initial screen (not shown) displayed on the UI 10. The information processing device 100 performs the login process based on the input user identification information and password.

[0097] (Step S2-8) The natural language search UI 18 accepts conditions for creating a classification that are input as a prompt on the Chat input screen of the LLM (not shown). The natural language search agent 64 acquires the user identification information and the conditions for creating a classification input as a prompt from the natural language search UI 18 , and outputs them to the pseudo metadata search agent 62 . The pseudo-metadata search agent 62 acquires user identification information and conditions for creating a classification input as a prompt from the natural language search agent 64. The pseudo-metadata search agent 62 outputs the acquired user identification information and prompt to the LLM (not shown). The LLM (not shown) acquires the user identification information and prompt from the pseudo-metadata search agent 62, references a (T+R+M+N)-dimensional vector space in which multiple contents stored in the content vector space storage unit 86 are embedded, and creates a content subspace that matches the conditions of the user U.

[0098] (Step S3-8) The vector space generation agent 84 creates and completes a classification in the content subspace that matches the conditions of the user U created by the LLM (not shown). Fig. 15 is a diagram showing an example of a content classification screen. As shown in Fig. 15, the example of the content classification screen includes a section displaying major categories and a section displaying medium categories. Furthermore, the example of the content classification screen includes an update button UPBU.

[0099] In the section displaying the major categories, one or more folders are displayed. In the example shown in Fig. 15, the "AAA" folder, the "BBB" folder, the "CCC" folder, ..., the "XXX" folder, ..., and the "KKK" folder are displayed. In the section displaying the intermediate categories, one or more folders included in the folder designated by the user U, out of the one or more folders displayed in the section displaying the major categories, are displayed. In the example shown in Fig. 15, the "PPP" folder, the "QQQ" folder, the "RRR" folder, ..., the "SSS" folder, and the "TTT" folder included in the "XXX" folder are displayed. Furthermore, by pressing the update button UPBU, the maximum number of layers m that can be set by the user U, the average number of contents n1 included in the folder at the bottom layer, and the number of folders k included in the i-th layer from the bottom layer are updated on the classification structure setting screen shown in Fig. 7. i This makes it possible to correct the above.

[0100] <Configuration example> As one configuration example, the information processing device includes a reception unit that receives content, a content processing unit that assigns information identifying a physical address to the content received by the reception unit, stores the content in a content pool, and virtually classifies the content stored in the content pool, and a display control unit that displays the results of the virtual content classification by the content processing unit on a display unit. By configuring the information processing device in this way, the content stored in the content pool can be virtually classified and displayed on the display unit, thereby reducing the effort required for content management.

[0101] In one configuration example, the content processing unit searches for content stored in a content pool, and the display control unit causes the display unit to display the results of the content search by the content processing unit. With this configuration, the information processing device can display content stored in a content pool on the display unit. For example, a virtual classification display UI may be created and displayed for each organization or user. When a user specifies content on the virtual classification display UI, the information processing device can display the specified content on the display unit. This reduces the effort required for content management. Here, a URI (Uniform Resource Identifier) ​​may be specified instead of or together with a physical address, causing the content corresponding to the specified URI to be displayed on the display unit.

[0102] In one configuration example, the content processing unit further includes a content pool crawler that receives information identifying physical addresses assigned to content stored in a content pool, a metadata generation unit that acquires the physical addresses from the content pool crawler, acquires information identifying the content from the content pool based on the acquired physical addresses, and generates metadata based on the acquired information identifying the content, a vector space generation unit that generates an N-dimensional vector space (N is an integer greater than 1), and an embedding processing unit that embeds content in the N-dimensional vector space generated by the vector space generation unit based on the metadata. The vector space generation unit generates k clusters (k is an integer greater than 1) from the multiple content pieces embedded in the N-dimensional vector space (N is an integer greater than 1). With this configuration, the information processing device can generate an N-dimensional vector space as metadata that represents the characteristics of the content. Furthermore, the information processing device can generate k clusters from multiple pieces of content embedded in the generated N-dimensional vector space. Because multiple pieces of content can be classified into k clusters, the effort required for content management can be reduced.

[0103] As one configuration example, the vector space generation unit derives inter-content distances between multiple pieces of content embedded in an N-dimensional vector space, and generates k clusters from the multiple pieces of content based on the derived inter-content distances. With this configuration, the information processing device can generate k clusters based on the inter-content distances between multiple pieces of content embedded in an N-dimensional vector space. Because multiple pieces of content can be classified into k clusters, the effort required for content management can be reduced.

[0104] In one configuration example, the vector space generation unit derives the distance between the contents further based on the weight vector. By configuring the information processing device in this manner, the information processing device can derive the distance between contents based additionally on the weight vector, thereby averaging the variation between each attribute to derive the distance between contents, or increasing the weight for an attribute whose influence on the distance between contents is desired to be increased.

[0105] In one configuration example, the content processing unit further includes a hierarchical structure generating unit that generates a hierarchical structure based on the distance between the centers of gravity of the k clusters. With this configuration, the information processing device can classify content based on the distance between the centers of gravity of the k clusters, thereby reducing the effort required to manage content.

[0106] As one configuration example, the vector space generation unit creates M (M is an integer greater than 1) dimensional vector data indicating the attributes of the content creator, generates an M dimensional vector space, and the embedding processing unit embeds the content in the N+M dimensional vector space. With this configuration, the information processing device can create M-dimensional vector data indicating the attributes of the content creator and generate an N+M-dimensional vector space. Furthermore, the information processing device can generate multiple clusters based on the inter-content distances between multiple pieces of content embedded in the generated N+N-dimensional vector space. Since multiple pieces of content can be classified into multiple clusters, the effort required for content management can be reduced.

[0107] As one configuration example, the vector space generation unit creates R-dimensional (R is an integer greater than 1) vector data based on the management rule, and generates an R-dimensional vector space. The embedding unit embeds the content in the N+M+R-dimensional vector space. With this configuration, the information processing device can create R-dimensional vector data based on the management rules and generate an N+M+R-dimensional vector space. Furthermore, the information processing device can generate multiple clusters based on the inter-content distances between multiple pieces of content embedded in the generated N+N+R-dimensional vector space. Since multiple pieces of content can be classified into multiple clusters, the effort required for content management can be reduced.

[0108] As an example configuration, the vector space generation unit creates T-dimensional vector data (T is an integer greater than 1) that indicates the time axis and attributes, and generates a T-dimensional vector space. The embedding unit embeds the content in the N+M+R+T-dimensional vector space. With this configuration, the information processing device can create T-dimensional vector data indicating the time axis and attributes, and generate an N+M+R+T-dimensional vector space. Furthermore, the information processing device can generate multiple clusters based on the inter-content distances between multiple pieces of content embedded in the generated N+M+R+T-dimensional vector space. Since multiple pieces of content can be classified into multiple clusters, the effort required for content management can be reduced.

[0109] As one configuration example, the vector space generation unit creates U-dimensional (U is an integer greater than 1) vector data indicating user attributes, generates a U-dimensional vector space, and the embedding processing unit embeds the content into the U+N+M+R+T-dimensional vector space. With this configuration, the information processing device can create U-dimensional vector data indicating user attributes and generate a U+N+M+R+T-dimensional vector space. Furthermore, the information processing device can generate multiple clusters based on the inter-content distances between multiple pieces of content embedded in the generated U+N+M+R+T-dimensional vector space. Because multiple pieces of content can be classified into multiple clusters, the effort required for content management can be reduced.

[0110] In one configuration example, the content processing unit generates pseudo-metadata based on a search prompt and user attributes, and further includes a pseudo-metadata search unit that searches for content based on the created pseudo-metadata, and the display control unit displays the content searched by the pseudo-metadata search unit on the display unit. With this configuration, the information processing device can generate pseudo metadata based on the search prompt and the user attributes, and can search for content based on the generated pseudo metadata.

[0111] In one configuration example, the vector space generation unit derives the similarity between the contents based on the distance between the contents, and the display control unit causes the display unit to display the similarity derived by the vector space generation unit. With this configuration, the information processing device can inform the user of the degree of similarity between a predetermined piece of content and another piece of content by notifying the user of the degree of similarity. Here, the distance between pieces of content can be measured by cosine similarity, Manhattan distance, Mahalanobis distance, Perceptual Distance, tree distance, or the like.

[0112] In one configuration example, the vector space generation unit derives inter-content distances between a plurality of pieces of content, further based on the content search results by the pseudo-metadata search unit. With this configuration, the information processing device can generate k clusters based on the inter-content distances between multiple pieces of content, further based on the content search results. Being able to classify multiple pieces of content into k clusters reduces the effort required for content management. Here, the method for classifying multiple pieces of content into k clusters does not depend on how the distance is defined. In addition to methods based on the k-nearest neighbor method, other clustering methods can also be applied.

[0113] As one configuration example, the hierarchical structure generation unit analyzes metadata for the content contained in each of the k clusters, and creates information indicating classification for recommending folders based on the results of the metadata analysis. By configuring in this manner, the information processing device can analyze the metadata for the content contained in each of the k clusters, and can create information indicating classifications for recommending folders based on the results of the metadata analysis.

[0114] As one configuration example, if there is missing information in the metadata generated based on the content, the metadata generation unit creates information for inquiring about the missing information, obtains information entered in response to the created information for inquiring, and completes the metadata. With this configuration, the information processing device can inquire of the user U about missing metadata, and can generate a vector space further based on the metadata acquired through the inquiry.

[0115] As one configuration example, the hierarchical structure generation unit generates a hierarchical structure based on the maximum number of levels, the average number of contents included in a folder in the lowest level, and the number of folders included in any level from the lowest level. With this configuration, the information processing device can generate a hierarchical structure based on the maximum number of layers set by the user U, the average number of contents contained in the folders in the lowest layer, and the number of folders contained in any layer from the lowest layer.

[0116] In one configuration example, the pseudo-metadata search unit derives content characteristics based on the content search results, and the display control unit displays information indicating the content characteristics derived by the pseudo-metadata search unit on the display unit. With this configuration, the information processing device can derive the features of the content based on the search results for the content and display them on the display unit, thereby allowing the user U to recognize the features of the content.

[0117] In one configuration example, the pseudo-metadata search unit derives questions for narrowing down the content based on the characteristics of the derived content, and the display control unit displays information indicating the questions for narrowing down the content derived by the pseudo-metadata search unit on the display unit. By configuring the information processing device in this manner, the information processing device can derive questions to narrow down the content based on the characteristics of the content and display them on the display unit, thereby enabling the optimal content to be selected based on the answers to the questions.

[0118] In one configuration example, if there is missing information in the user attributes, the pseudo-metadata search unit creates information for inquiring about the missing information. With this configuration, the information processing device can inquire of the user U about the missing user attributes, and can search for content further based on the user attributes acquired through the inquiry.

[0119] Although an embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and design changes and the like are also included within the scope that does not deviate from the gist of the present invention. Furthermore, a computer program for realizing the functions of the information processing device 100 described above may be recorded on a computer-readable recording medium, and the program recorded on this recording medium may be read and executed by a computer system. Note that the "computer system" referred to here may include hardware such as an OS and peripheral devices.

[0120] In addition, "computer-readable recording medium" refers to writable non-volatile memory such as a flexible disk, optical magnetic disk, ROM, or flash memory, portable media such as a DVD (Digital Versatile Disk), or a storage device such as a hard disk built into a computer system.

[0121] Furthermore, the term "computer-readable recording medium" also includes those that retain a program for a certain period of time, such as volatile memory (e.g., DRAM (Dynamic Random Access Memory)) within a computer system that serves as a server or client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line.

[0122] The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by transmission waves in the transmission medium. Here, the "transmission medium" that transmits the program refers to a medium that has the function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.

[0123] The program may also be a program for implementing some of the functions described above, or may be a so-called differential file (differential program) that can implement the functions described above in combination with a program already stored in the computer system. [Explanation of symbols]

[0124] 100...information processing device, 10...UI, 12...account registration / update UI, 14...content pool UI, 16...categorization display UI, 18...natural language search UI, 20...account registration / update processing unit, 30...content pool, 40...virtual content pool construction processing unit, 42...content pool crawler, 44...metadata generation agent, 50...hierarchical structure generation agent, 62...pseudo metadata search agent, 64...natural language search agent, 66...display control unit, 80...virtual content pool, 81...user attribute DB, 82...metadata DB, 83...usage log DB, 84...vector space generation agent, 85...embedding processing unit, 86...content vector space storage unit

Claims

1. a reception unit that receives content; a content processing unit that acquires a plurality of pieces of content from a content pool in which the plurality of pieces of content received by the receiving unit are assigned information specifying a physical address and stored, based on each of the pieces of content specifying a physical address, generates metadata including information corresponding to user attribute information based on each of the acquired pieces of content, generates a vector space expressing the characteristics of the content based on the generated pieces of metadata and user attribute information of a user viewing any of the plurality of pieces of content, and virtually classifies the content by generating a plurality of clusters from the plurality of pieces of content based on the generated vector space; a display control unit that causes a display unit to display the result of the virtual classification of the content by the content processing unit; An information processing device comprising:

2. the content processing unit searches for content stored in the content pool based on a search request received via a user interface; The information processing device according to claim 1 , wherein the display control unit causes a display unit to display a result of the content search performed by the content processing unit.

3. The content processing unit a content pool crawler that receives information identifying a physical address assigned to content stored in the content pool; a metadata generation unit that acquires information specifying a physical address from the content pool crawler, acquires content from the content pool based on the acquired information specifying the physical address, and generates metadata based on the acquired content; a vector space generation unit that generates an N-dimensional (N is an integer greater than 1) vector space as metadata that expresses features of the content, based on the plurality of metadata and the user attribute information of a user who browses any of the plurality of content items; an embedding processing unit that embeds content into the N-dimensional vector space generated by the vector space generation unit based on metadata; Equipped with the vector space generation unit generates k (k is an integer greater than 1) clusters from a plurality of contents embedded in an N-dimensional vector space (N is an integer greater than 1); N is the number of attribute items to be vectorized among the plurality of metadata. The information processing device according to claim 2 .

4. The information processing device according to claim 3 , wherein the vector space generation unit derives inter-content distances between the plurality of pieces of content embedded in the N-dimensional vector space, and generates k clusters from the plurality of pieces of content based on the derived inter-content distances.

5. The information processing device according to claim 4 , wherein the vector space generation unit derives the distance between the contents further based on the weight vector.

6. The content processing unit a hierarchical structure generating unit that generates a hierarchical structure based on the distances between the centers of gravity of the k clusters; The information processing device according to claim 3 , further comprising:

7. the vector space generation unit generates M-dimensional vector data (M is an integer greater than 1) indicating attributes of a content creator, and generates an M-dimensional vector space; The information processing device according to claim 3 , wherein the embedding processing unit embeds the content in an N+M-dimensional vector space.

8. the vector space generation unit generates R-dimensional vector data (R is an integer greater than 1) based on a management rule, and generates an R-dimensional vector space; The information processing device according to claim 7 , wherein the embedding processing unit embeds the content in an N+M+R dimensional vector space.

9. the vector space generation unit creates T-dimensional vector data (T is an integer greater than 1) indicating a time axis and attributes, and generates a T-dimensional vector space; The information processing device according to claim 8 , wherein the embedding processing unit embeds the content in an N+M+R+T dimensional vector space.

10. The content processing unit a pseudo-metadata search unit that generates pseudo-metadata based on the search prompt and user attributes and searches for content based on the generated pseudo-metadata; Furthermore, The information processing device according to claim 3 , wherein the display control unit causes the content searched for by the pseudo metadata search unit to be displayed on a display unit.

11. the vector space generation unit derives a similarity between the contents based on a distance between the contents; The information processing device according to claim 4 , wherein the display control unit causes a display unit to display the similarity derived by the vector space generation unit.

12. the vector space generation unit derives a content distance between the content searched by the pseudo-metadata search unit and other content; The information processing device according to claim 10 , wherein the display control unit causes the display unit to display, as a search result, other content that has a short inter-content distance from the content searched for by the pseudo metadata search unit.

13. 7. The information processing device according to claim 6, wherein the hierarchical structure generation unit analyzes metadata for the content included in each of the k clusters, and creates information indicating a classification for recommending folders corresponding to each of the k clusters based on a result of the analysis of the metadata.

14. 4. The information processing device according to claim 3, wherein, when the metadata generated based on the content contains missing information, the metadata generation unit creates information for inquiring about the missing information, acquires information input in response to the created information for inquiring, and completes the metadata.

15. 7. The information processing device of claim 6, wherein the hierarchical structure generation unit calculates the number of folders to be included in each layer based on parameters of the hierarchical structure to be generated, such as the maximum number of layers, the maximum number of layers corresponding to the cluster in the lowest layer, the average number of contents included in a folder in the lowest layer, and the number of folders included in any layer from the lowest layer.

16. the pseudo-metadata search unit derives content features based on content search results; The information processing device according to claim 10 , wherein the display control unit causes the display unit to display the information indicating the characteristics of the content derived by the pseudo metadata search unit.

17. the pseudo-metadata search unit derives a query for narrowing down the content based on the derived content characteristics; The information processing device according to claim 16 , wherein the display control unit causes the display unit to display information indicating a question for narrowing down the content derived by the pseudo metadata search unit.

18. The information processing apparatus according to claim 10 , wherein, when there is insufficient information in the user attributes, the pseudo-metadata search unit creates information for inquiring about the insufficient information.

19. 1. A computer-implemented information processing method, comprising: Accepts multiple contents, acquiring a plurality of pieces of content from a content pool in which information specifying a physical address is assigned to each of the plurality of received pieces of content and stored, based on each of the pieces of information specifying a physical address; generating metadata including information corresponding to user attribute information based on each of the acquired plurality of pieces of content; and generating a vector space expressing the characteristics of the content based on the generated plurality of pieces of metadata and user attribute information of a user viewing any of the plurality of pieces of content; generating a plurality of clusters from a plurality of pieces of content based on the generated vector space, thereby virtually classifying the pieces of content; The results of the virtual classification of the content are displayed on a display unit. Information processing methods.

20. On the computer, Accepts multiple contents, acquiring a plurality of pieces of content from a content pool in which information specifying a physical address is assigned to each of the plurality of received pieces of content and stored, based on each of the pieces of information specifying a physical address; generating metadata including information corresponding to user attribute information based on each of the acquired plurality of pieces of content; and generating a vector space expressing the characteristics of the content based on the generated plurality of pieces of metadata and user attribute information of a user viewing any of the plurality of pieces of content; generating a plurality of clusters from a plurality of pieces of content based on the generated vector space, thereby virtually classifying the pieces of content; The results of the virtual classification of the content are displayed on a display unit. program.

Images