AI large model-based intelligent management system for delivery of known products

Abstract

The invention relates to the technical field of delivery data processing, and discloses an AI large model-based intelligent management system for delivery of known products, and the system comprises an attribute acquisition module, a temporary scheme generation module, a key feature element identification module and a type attribute identification module, the optimization module is used for dynamically distributing a task priority of the delivery request based on the temporary correction scheme and synchronously optimizing an execution path of the delivery request based on the task priority; the conflict diffusion module is used for gathering task conflict traces of the delivery request based on the optimized execution path and the logic chain of the task priority, and constructing a conflict diffusion topological graph of the delivery request through the task conflict traces; the delivery module deploys an optimal delivery path based on the resolution result, and can solve the problem that in the prior art, the high-complexity and high-timeliness delivery requirements of the known products cannot be met.

Description

Technical Field

[0001] This invention relates to the field of delivery data processing technology, and in particular to an intelligent management system for intellectual property delivery based on AI big data models. Background Technology

[0002] Existing technologies often only enable basic document archiving and process recording, failing to dynamically identify potential conflicts between key features of delivery requests and document type attributes. This frequently leads to delivery delays caused by issues such as format mismatches and missing elements. Furthermore, task priority allocation lacks quantitative basis and relies heavily on experience-based judgment, easily resulting in insufficient or redundant resource allocation for important tasks. Fixed and rigid execution paths also make it difficult to dynamically adjust to real-time conflicts, severely impacting delivery efficiency and accuracy.

[0003] Furthermore, existing technologies for handling task conflicts mostly remain at the level of single-point resolution, lacking the ability to systematically analyze and predict conflict escalation paths. When task node conflicts occur, it is impossible to trace the source of the conflict and its potential scope of impact through topology, leading to repeated conflicts or their spread. Simultaneously, the generation of optimal delivery paths lacks deep integration of conflict resolution results and struggles to dynamically optimize paths by combining resource distribution and task priorities. This often results in resource waste and process redundancy during intellectual property delivery, failing to meet the demands of highly complex and time-sensitive intellectual property delivery. Summary of the Invention

[0004] This invention provides an intelligent management system for intellectual property delivery based on AI large models to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides an intelligent management system for intellectual property delivery based on an AI large-scale model, the system comprising:

[0006] Attribute acquisition module: used to receive intellectual property delivery requests and document resources, and synchronously acquire the key feature elements of the delivery request and the type attributes of the document resources;

[0007] Temporary solution generation module: used to dynamically identify potential interaction conflicts between the key feature elements and the type attributes, and generate a temporary correction solution for the delivery request based on the potential interaction conflicts;

[0008] Optimization module: used to dynamically allocate the task priority of the delivery request based on the temporary correction scheme, and simultaneously optimize the execution path of the delivery request based on the task priority;

[0009] Conflict propagation module: used to aggregate task conflict traces of the delivery request based on the logical chain of the optimized execution path and the task priority, and to construct a conflict propagation topology graph of the delivery request through the task conflict traces;

[0010] Delivery module: used to resolve the task conflict traces based on the conflict diffusion topology graph, and deploy the optimal delivery path based on the resolution result.

[0011] Preferably, the step of simultaneously acquiring the key feature elements of the delivery request and the type attribute of the document resource includes:

[0012] The content features of the delivery request are analyzed, the key feature elements of the delivery request are extracted from the content features, and the type attributes of the document resource are reviewed simultaneously.

[0013] Preferably, the dynamic identification of potential interaction conflicts between the key feature elements and the type attributes includes:

[0014] Construct the rule conflict coefficient matrix of the delivery request based on the entropy difference results of the key feature elements and the type attributes;

[0015] Extend the structured causal chain topology along the rule conflict coefficient matrix;

[0016] The results of the topology extension are used to identify potential conflict propagation paths in the delivery request;

[0017] Verify the pattern anomaly of the conflict propagation path, locate the endpoint of the abnormal propagation path of the delivery request based on the pattern anomaly, and identify the endpoint of the abnormal propagation path as a potential interaction conflict of the delivery request.

[0018] Preferably, the provisional correction scheme for generating the delivery request based on the potential interaction conflict includes:

[0019] Based on the potential interaction conflicts, identify the conflict points between the key feature elements and the type attributes;

[0020] By correcting the historical baseline data of the delivery request based on the conflict points, a preliminary correction proposal for the delivery request is generated;

[0021] Based on the aforementioned preliminary revised proposal, the environmental parameters were subjected to feasibility simulation verification to obtain the conflict simulation screening results.

[0022] The conflict simulation screening is integrated into a temporary correction scheme.

[0023] Preferably, the task priority of the delivery request is dynamically allocated based on the temporary correction scheme, wherein the calculation formula for dynamic allocation is: in: As a task priority, Let be the conflict entropy decay function. For the first The element deviation degree at each point of conflict For the first The path coupling coefficient at each conflict point As a dynamic correction factor, The maturity level of the temporary revised scheme is... For time-sensitive multipliers, and For learnable parameters, As the knowledge intensity decay factor, The total number of conflict points. For indexing.

[0024] Preferably, the step of simultaneously optimizing the execution path of the delivery request based on the task priority includes:

[0025] Based on the task priority and the temporary correction scheme, an intermediate optimized path for the delivery request is generated;

[0026] Verify the conflict compatibility of the intermediate optimization path;

[0027] The intermediate optimization path that passes the verification is output as the optimized execution path.

[0028] Preferably, the logical chain aggregation of the delivery request task conflict traces based on the optimized execution path and the task priority includes:

[0029] Analyze the sequence of task nodes in the optimized execution path and calculate the sequence offset of the task nodes in the sequence;

[0030] When the sequence offset exceeds a preset conflict detection threshold, the corresponding task node is marked as a potential conflict trigger point;

[0031] By integrating the interaction chains of the potential conflict trigger points through the logical chain, the task conflict trace of the delivery request is obtained.

[0032] Preferably, constructing the conflict propagation topology map of the delivery request through the task conflict traces includes:

[0033] Analyze the conflict intensity of the task conflict traces;

[0034] Identify areas where the conflict intensity exceeds a preset propagation threshold, and mark the areas exceeding the threshold as conflict outbreak areas;

[0035] Connect the conflict outbreak areas and simultaneously extend the conflict range along the task dependency direction of the delivery request to obtain the initial diffusion path of the delivery request;

[0036] The integrity of the preliminary diffusion path is verified, and the verified preliminary diffusion paths are pieced together to form a conflict diffusion topology map.

[0037] Preferably, the step of resolving the task conflict traces based on the conflict diffusion topology graph includes:

[0038] Analyze the distribution of conflict hotspots in the aforementioned conflict diffusion topology map;

[0039] Based on the task priority, the order of the diffusion paths of the conflict hotspots in the conflict hotspot distribution is adjusted, and the result of the adjustment is used as a preliminary resolution plan.

[0040] Verify the completeness of the task conflict traces in the preliminary resolution plan. When the completeness of the resolution exceeds a preset stability threshold, integrate and output the resolved task conflict traces.

[0041] Preferably, the step of deploying the optimal delivery path based on the resolution result includes:

[0042] The resource nodes in the resolved task conflict traces are aggregated into key resource nodes:

[0043] Based on the task priority, the key resource nodes are connected, and a preliminary delivery path sequence is generated along the gradient guidance trajectory of the task priority;

[0044] Verify the execution efficiency integrity of the initial delivery path sequence. When the efficiency gain value meets the dynamic deployment threshold, reconstruct and output the optimal delivery path.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. By leveraging the attribute acquisition module and the temporary solution generation module, key feature elements and type attributes can be deeply extracted from delivery requests and document resources. A rule conflict coefficient matrix is ​​constructed using entropy difference results to accurately identify potential interaction conflicts and generate targeted temporary correction solutions. This avoids the problems of delayed conflict detection and one-sided processing in traditional manual review, significantly reducing the risk of delivery delays caused by conflicts. Simultaneously, task priorities are dynamically allocated based on quantitative parameters such as element deviation and path coupling coefficient at conflict points, ensuring resources are tilted towards high-priority tasks. This solves the subjectivity and resource misallocation problems of traditional experience-based priority allocation.

[0047] 2. By constructing a conflict diffusion topology map through the conflict diffusion module, global tracking of task conflict traces and prediction of diffusion paths are achieved. Combined with the orderly resolution of conflict hotspots and the aggregation of key resource nodes by the delivery module, the optimal delivery path is ultimately generated. This process not only upgrades conflict resolution from single-point to systemic deconstruction, but also significantly improves the smoothness of the delivery process and resource utilization efficiency through dynamic optimization and efficiency verification of execution paths. Compared to traditional fixed processes, the system can adapt to the complex needs of different delivery scenarios, significantly shortening the delivery cycle while ensuring delivery accuracy, and providing an intelligent and refined solution for intellectual property delivery management. Attached Figure Description

[0048] Figure 1 A functional module diagram of an AI-based intelligent management system for intellectual property delivery, provided in an embodiment of the present invention; Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments belong to some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “said” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0051] Depending on the context, the word "if" or "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0052] This application provides an intelligent management system for intellectual property delivery based on an AI-based large-scale model. The executing entity of this intelligent management system includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the intelligent management system for intellectual property delivery based on an AI-based large-scale model can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks, and big data and artificial intelligence platforms.

[0053] like Figure 1 The diagram shown is a functional module diagram of the intelligent management system for intellectual property delivery based on an AI big model, according to the present invention.

[0054] The AI-based large-scale model-based intelligent management system for intellectual property delivery 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the AI-based large-scale model-based intelligent management system 100 may include an attribute acquisition module 101, a temporary solution generation module 102, an optimization module 103, a conflict diffusion module 104, and a delivery module 105. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0055] In this embodiment, the functions of each module / unit are as follows:

[0056] Attribute acquisition module 101: used to receive intellectual property delivery requests and document resources, and synchronously acquire the key feature elements of the delivery request and the type attributes of the document resources.

[0057] In this embodiment, the step of simultaneously acquiring the key feature elements of the delivery request and the type attribute of the document resource includes:

[0058] The content features of the delivery request are analyzed, the key feature elements of the delivery request are extracted from the content features, and the type attributes of the document resource are reviewed simultaneously.

[0059] Specifically, a delivery request refers to a user's specific instruction regarding the delivery of intellectual property rights, including information such as the type of intellectual property rights to be delivered, the recipient, the delivery time limit, and the delivery standards.

[0060] Document resources are various documents and materials related to intellectual property to be delivered, such as patent applications, trademark registration certificates, original manuscripts of works, and proof of rights. These documents may be stored in the database or cloud storage connected to the system in the form of electronic documents.

[0061] Key feature elements are the core information extracted from the content features of the delivery request, such as the urgency of delivery, the ownership status of intellectual property rights, the technical field involved, and the quantity and scale of delivery. These elements are important bases for subsequent processing and decision-making.

[0062] Type attributes are the category characteristics inherent in a document resource, including the document's format, content type, creation time, version information, etc.

[0063] In detail, the receiving module actively receives intellectual property delivery requests sent from user terminals through the system interface, and simultaneously retrieves the document resources associated with the delivery request from the document repository. This action acts as an information entry point, bringing external needs and related materials into the system's processing scope.

[0064] Parsing involves analyzing the content of received delivery requests. Using AI technologies such as natural language processing, it breaks down the text of the delivery request, identifies key information points, and extracts key feature elements. For example, from a request to urgently deliver the authorization documents for a certain invention patent to Company A within three days, key feature elements such as urgency, intellectual property type, and recipient can be extracted.

[0065] The review process involves extracting key features of the delivery request while simultaneously determining and examining the type and attributes of the acquired document resources. This includes checking whether the document format meets system processing requirements, whether the content is relevant to the delivery request, and whether there are version errors, ensuring the validity and applicability of the document resources. For example, reviewing a patent drawing document might determine that its format is PDF and it is the latest version, thus qualifying it as the required drawing material for the patent delivery.

[0066] Temporary solution generation module 102: used to dynamically identify potential interaction conflicts between the key feature elements and the type attributes, and generate a temporary correction solution for the delivery request based on the potential interaction conflicts.

[0067] In this embodiment, the dynamic identification of potential interaction conflicts between the key feature elements and the type attributes includes:

[0068] Construct the rule conflict coefficient matrix of the delivery request based on the entropy difference results of the key feature elements and the type attributes;

[0069] Extend the structured causal chain topology along the rule conflict coefficient matrix;

[0070] The results of the topology extension are used to identify potential conflict propagation paths in the delivery request;

[0071] Verify the pattern anomaly of the conflict propagation path, locate the endpoint of the abnormal propagation path of the delivery request based on the pattern anomaly, and identify the endpoint of the abnormal propagation path as a potential interaction conflict of the delivery request.

[0072] Specifically, key feature elements are the core information derived from the delivery request, such as the time limit requirements for intellectual property delivery, the delivery recipient, and the ownership status of the intellectual property. These are the key dimensions for defining delivery requirements.

[0073] Type attributes are inherent characteristics of document resources or process steps, such as document format, process node type, and intellectual property type, which describe the inherent characteristics of the elements involved in delivery.

[0074] Entropy difference results are a measure of the degree of disorder and uncertainty in a system. Here, by calculating the uncertainty differences between key feature elements and type attributes in dimensions such as information matching degree and process adaptability, the numerical results reflect the potential incompatibility and conflict trends between the two. The greater the difference, the higher the probability of conflict.

[0075] The rule conflict coefficient matrix is ​​a matrix that quantifies the probability of conflict corresponding to different combinations of key feature elements and type attributes. The rows and columns can correspond to the categories of key feature elements and type attributes, respectively, and the matrix element values ​​represent the conflict coefficients of the two, which are used to intuitively show the multi-dimensional conflict distribution.

[0076] In detail, the process begins by retrieving key feature elements and type attribute data stored in the database. Then, using a conflict analysis algorithm trained on a large AI model, the entropy difference between each set of key feature elements and type attributes is calculated. Based on this entropy difference, a pre-defined matrix construction rule is followed, such as row-first traversal of feature elements and corresponding column attribute categories. This process fills in matrix element values, generates a rule-based conflict coefficient matrix, and constructs a conflict probability map, thus presenting the scattered conflict trend data in a structured manner.

[0077] Specifically, a structured causal chain is a chain of causal relationships formed based on business logic and rules in the intellectual property delivery process. For example, a document format error → failure to pass the initial system review → delivery delay is a potential logical path for the propagation of conflicts and problems. In the matrix, it is reflected as the relationship between elements formed by the association of processes and rules.

[0078] Topological extension, based on topological concepts, focuses on expanding and traversing the relationships between elements rather than specific numerical values. In a matrix, this means following the causal chain, such as extending from conflict nodes between document format type attributes and key delivery timeframes to conflicts associated with nodes in the initial review process, uncovering potential conflict propagation and related nodes, and expanding the scope of conflict analysis.

[0079] In detail, based on the existing causal relationship markers of the elements in the matrix, and trained by business rules and historical conflict patterns, starting from the conflict nodes of the rule conflict coefficient matrix, the related matrix elements are traversed sequentially along the logical direction of the structured causal chain, such as the business flow order of document processing → process review → delivery output. This expands the conflict analysis from a single feature-attribute combination to all elements involved in the entire relationship chain, like digging out the possible path of conflict spread by following clues, and connecting isolated conflict points into a potential conflict propagation network.

[0080] Specifically, the result of topological extension is the set of conflict-related nodes and the network of relationships obtained after the previous step of extending along the causal chain. It includes multiple links that the conflict may go through and the combination of features and attributes involved, and is a set of conflict propagation clue maps.

[0081] The conflict propagation path is the specific route in the intellectual property delivery process from the initial trigger point, such as incompatibility between a feature and an attribute, along business logic and rules, to subsequent stages. For example, a conflict between delivery deadline requirements and document review cycles can lead to review timeouts → delivery delays → default risks. This is a conflict propagation path that needs to be identified and controlled.

[0082] In detail, the AI ​​algorithm performs path mining on the interconnected network obtained from the topology extension. Using graph theory algorithms, such as shortest path and depth-first search, it identifies the complete routes from which conflicts may propagate from the interconnected nodes. For example, in the node-connected network formed by the topology extension, it marks the complete path from the initial conflicting element, such as the conflict between expedited delivery features and complex document attributes, through review process nodes, communication feedback nodes, etc., to the node that ultimately affects the delivery result. This clearly outlines the potential conflict transmission routes, providing direction for subsequent investigation and correction.

[0083] Specifically, pattern anomaly measures the degree of deviation between the conflict propagation path and the normal business process pattern, which is formed by training a conventional path model based on historical data and standard processes. The more the path deviates from the conventional process logic and the more likely it is to cause serious problems, such as delivery failure or major risks, the higher the pattern anomaly.

[0084] Abnormal propagation path endpoints are the nodes in the conflict propagation path that deviate most significantly from the normal pattern and have a critical impact on the delivery result. They may be the source of the conflict, such as unreasonable delivery time limits, or they may be the final key impact point of the conflict transmission, such as delivery delays that lead to breach of contract. These are the core conflict locations that need to be corrected and controlled.

[0085] Potential interaction conflicts are the points of interaction contradictions that may cause problems between key feature elements and type attributes in the intellectual property delivery process, identified through the above analysis. These are the targets for generating temporary correction solutions.

[0086] In detail, a normal pattern model trained using historical intellectual property delivery process data is used to compare the identified conflict propagation paths with this model. The algorithm calculates the path pattern anomaly, such as counting the number and degree of differences between path nodes and normal nodes, and combines this with risk weight calculations to filter out paths with anomalies exceeding a threshold. Then, endpoint localization is performed on these anomaly paths to identify the most critical and most unusual starting or ending nodes. For example, in a delivery timeframe conflict propagation path, the source endpoint of unreasonably shortening the review period is located and marked as a potential interaction conflict in the delivery request, serving as the core issue to be addressed in the temporary correction plan.

[0087] In this embodiment, the step of generating a temporary correction scheme for the delivery request based on the potential interaction conflict includes:

[0088] Based on the potential interaction conflicts, identify the conflict points between the key feature elements and the type attributes;

[0089] By correcting the historical baseline data of the delivery request based on the conflict points, a preliminary correction proposal for the delivery request is generated;

[0090] Based on the aforementioned preliminary revised proposal, the environmental parameters were subjected to feasibility simulation verification to obtain the conflict simulation screening results.

[0091] The conflict simulation screening is integrated into a temporary correction scheme.

[0092] Specifically, potential interaction conflicts are key feature elements in the intellectual property delivery process identified in the early stages of the system, such as delivery time limits and ownership requirements, and type attributes, such as document format and process node attributes, that may cause problems. For example, the incompatibility between the expedited delivery feature and the multi-level review attribute of the document is the source of problems in subsequent analysis.

[0093] Key feature elements are the core requirement dimensions parsed from the delivery request, such as delivery time, delivery object, and intellectual property ownership status, which define the key information for delivery.

[0094] Type attributes are inherent characteristics of document resources, process steps, etc., such as document format (PDF / Word), process node type (preliminary review / final review), intellectual property type (patent / trademark), describing the inherent characteristics of the elements involved in delivery.

[0095] The point of conflict is the specific point of contradiction when key feature elements interact with type attributes. For example, the direct contradiction between delivery time limit requirements and document review cycle attributes is the target point to be corrected.

[0096] In detail, by retrieving previously marked potential interaction conflict data from the database, and combining it with stored key feature elements and type attribute information, and leveraging the semantic understanding and conflict correlation analysis capabilities of the AI ​​big data model, the specific contradictions in the interaction between the two are located. In the delivery requirement-process attribute correlation network, the nodes where feature A (urgent) collides with attribute B (slow review) are accurately marked as conflict points, which are then identified as clear targets for subsequent corrections.

[0097] Specifically, the conflict point is the specific contradiction identified in the previous step, such as a conflict between the delivery deadline and the review cycle, which is the focus of the corrective action.

[0098] Historical baseline data for delivery requests are standard data stored in the system for similar intellectual property delivery requests in the past, such as regular delivery time limits, process cycles, and resource configurations. It serves as a reference template for correction and includes various parameter thresholds and relationships for the normal delivery process.

[0099] The preliminary revision proposal is a preliminary delivery process modification plan formed after adjusting historical baseline data to address the conflict points. For example, it may modify the delivery time limit or optimize the review process nodes. It serves as the initial plan for subsequent verification.

[0100] In detail, historical baseline data matching the current delivery request is extracted from the historical database, such as past data on the same intellectual property type and similar delivery objects. Based on the conflict points, the baseline data is adjusted accordingly: if the conflict point is a short delivery deadline versus a long review cycle, the review process parameters in the historical baseline data are intelligently adjusted, such as compressing review steps, parallelizing review steps, or modifying the delivery deadline requirements. After adjustment, a preliminary revision proposal is generated according to a preset solution template, including process steps, parameter configurations, and resource allocation, forming a draft solution to resolve the conflict.

[0101] Specifically, the preliminary revision proposal is a revision scheme generated in the previous step that needs to be verified, such as the adjusted delivery process and parameters, and serves as the input object for simulation verification.

[0102] Environmental parameters are data on various influencing factors in the environment in which the intellectual property delivery process takes place, covering system resources, human resource allocation, external constraints, etc., simulating the conditional factors of real delivery scenarios.

[0103] Feasibility simulation verification uses the system's built-in simulation engine to simulate the process of the initial revised proposal running under actual environmental parameters, verifying whether the solution is feasible, and is a step to preview the implementation effect of the solution.

[0104] The conflict simulation screening results are the selected corrective solutions that are feasible under the environmental parameters after simulation verification, or the infeasible points and adjustment suggestions are marked, which serve as the basis for solution optimization.

[0105] In detail, the system invokes the simulation engine, first loading the preliminary revised proposal's workflow logic and parameter configuration. Simultaneously, it extracts real-time / historical environmental parameters from the environment database to construct the simulation runtime environment. The simulation engine then simulates the operation of each stage under environmental parameter constraints, following the proposal's defined workflow, such as document submission → review → feedback → delivery.

[0106] If the proposal shortens the review period, the simulation will calculate whether the existing review manpower can complete the task within that period and whether the server will experience lag when processing documents concurrently. After running, the simulation results will be output, marking feasible process segments and infeasible conflict points in the solution, forming conflict simulation screening results, and determining whether the initial proposal is successful.

[0107] Specifically, the conflict simulation screening result is a set of results obtained after simulation verification, containing information on the feasibility / infeasibility of the solution. It includes effective corrective measures that can be retained, as well as conflict points that need to be adjusted and suggestions.

[0108] The temporary correction plan is a complete plan formed by integrating simulation screening results and can be temporarily used to handle the current intellectual property delivery conflict. It clarifies the process adjustment, parameter configuration, resource scheduling and other contents, and directly guides the delivery execution.

[0109] In detail, the results of the conflict simulation screening were analyzed: feasible corrective measures were extracted, such as shortening the review cycle of parallel parts with sufficient resources; for infeasible points, solutions were supplemented based on simulation suggestions. Then, according to the standardized format of intellectual property delivery solutions, the screened effective content was integrated and pieced together to form a logically complete and executable temporary corrective solution, which was stored in the system solution library and pushed to relevant execution positions to guide the correction of the actual intellectual property delivery process.

[0110] Optimization module 103: used to dynamically allocate the task priority of the delivery request based on the temporary correction scheme, and simultaneously optimize the execution path of the delivery request based on the task priority.

[0111] In this embodiment, the task priority of the delivery request is dynamically allocated based on the temporary correction scheme, wherein the calculation formula for dynamic allocation is: in: As a task priority, Let be the conflict entropy decay function. For the first The element deviation degree at each point of conflict For the first The path coupling coefficient at each conflict point As a dynamic correction factor, The maturity level of the temporary revised scheme is... For time-sensitive multipliers, and For learnable parameters, As the knowledge intensity decay factor, The total number of conflict points. For indexing.

[0112] Specifically, The task priority is the final calculated output, representing the priority level of the intellectual property delivery task in the system scheduling. The higher the value, the higher the priority, and it is used to guide the system resource allocation order.

[0113] The conflict entropy decay function is used to quantify the reduction in uncertainty of conflict points as the process progresses and the solution is modified.

[0114] It is important to note that The calculation formula is: in: Let be the conflict entropy decay function. The entropy decay coefficient is derived from historical experience. The exponential decay characteristic demonstrates the mitigating effect of the correction scheme on conflict.

[0115] Specifically, For the first The element deviation at the first point of conflict measures the degree of deviation of the first conflict point. Within each point of conflict, the degree of deviation between key characteristic elements and type attributes. For example, the time difference between the delivery deadline requirement and the actual process cycle is calculated in the system database by comparing the conflict point data with standard parameters.

[0116] The calculation formula is: in: For the first The element deviation degree at each point of conflict For the first Key characteristic elements of each point of conflict For the first The type attribute of each conflict point As a weighting factor, This is derived from historical experience.

[0117] Specifically, For the first The path coupling coefficient of the first conflict point reflects the path coupling coefficient of the first conflict point. The degree of correlation between each conflict point and other links and conflict points in the delivery process path. In the physical environment, based on the topological modeling of the intellectual property delivery process by the system, the connection weight of the conflict point in the process network is calculated.

[0118] The calculation formula is: in: For the first The path coupling coefficient at each conflict point For the first The first point of conflict and the The distance between the conflict points.

[0119] Specifically, This is a dynamic correction factor, a coefficient that is dynamically adjusted based on the effect of temporary correction schemes on conflict points. During system operation, the AI ​​large model updates this factor in real time within the algorithm based on the feedback from scheme execution, reflecting the dynamic correction capability of the scheme.

[0120] The calculation formula is: in: As a dynamic correction factor, For the first The element deviation degree at each point of conflict For the first The path coupling coefficient at each conflict point This is the threshold constant.

[0121] Specifically, The maturity level of the temporary amendment scheme is an indicator that measures its completeness and feasibility. It is calculated based on system simulation verification and historical scheme comparison data, using multi-dimensional evaluation in the database, such as conflict resolution coverage and resource adaptability.

[0122] The time-sensitive multiplier reflects the impact of task delivery urgency on priority. It correlates with system timestamp data and delivery deadlines; the more urgent the time (e.g., approaching the delivery deadline), the larger this value, amplifying the task's priority.

[0123] and The learnable parameters are parameters trained on historical data and iteratively optimized using large AI models to adapt to the conflict handling patterns in different intellectual property delivery scenarios. When the server runs the model training algorithm, these two parameters are adjusted based on massive amounts of historical task data to make priority calculation more accurate.

[0124] The knowledge intensity decay factor takes into account how the effectiveness of historical experience and standard solutions diminishes over time and with changing scenarios in the intellectual property delivery process. The system relies on a knowledge graph and historical solution library stored in the database, combined with the differences between current and historical tasks, to calculate this decay factor, preventing outdated experience from affecting priority judgment.

[0125] Furthermore, The algorithm indicates that it is based on the index. traverse them one by one There are several conflict points. For each conflict point, The calculations are then performed and summed. The server's CPU processes each conflict point sequentially within the algorithm's loop.

[0126] It involves substituting the intermediate results into the conflict entropy decay function. Multiplied by the knowledge intensity decay factor and time-sensitive multiplier The server is running. Functional model, combined Historical experience decay setting, The time urgency data incorporates conflict attenuation and time impact into the results.

[0127] In this embodiment, the step of simultaneously optimizing the execution path of the delivery request based on the task priority includes:

[0128] Based on the task priority and the temporary correction scheme, an intermediate optimized path for the delivery request is generated;

[0129] Verify the conflict compatibility of the intermediate optimization path;

[0130] The intermediate optimization path that passes the verification is output as the optimized execution path.

[0131] Specifically, task priority is a numerical value representing the urgency and importance of intellectual property delivery tasks, such as high / medium / low priority, or a specific quantitative value, calculated in advance by the system. It serves as the basis for path optimization and determines the direction of resource allocation.

[0132] The temporary amendment plan is a preliminary process adjustment plan generated in response to intellectual property delivery conflicts. It is the basic framework for path optimization, clarifying the conflicts to be resolved and the direction of adjustment.

[0133] A delivery request is a user-initiated instruction to deliver intellectual property, and it is the target of path optimization. All optimizations revolve around satisfying this request.

[0134] The intermediate optimization path is a preliminary, unverified intellectual property delivery execution path that combines task priorities and temporary correction plans. For example, the adjusted process of document review → conflict correction → secondary review → delivery is a candidate solution for the formal execution path.

[0135] In detail, the task priority calculation results and temporary adjustment plan content are retrieved from the database. Guided by the delivery request, and based on priority (e.g., high-priority tasks are allocated computing power and human resources first), the process steps and resource scheduling in the temporary adjustment plan are adapted and adjusted. If the task priority is high, the algorithm will prioritize connecting key links and allocating expedited resources based on the temporary adjustment plan. Through a process topology modeling algorithm, the adjusted steps and resources are integrated to generate an intermediate optimized path, forming the delivery execution route.

[0136] Specifically, conflict compatibility refers to the degree of adaptation between the intermediate optimization path and various conflicts in the intellectual property delivery process, such as feature-attribute conflicts identified in the early stage and secondary conflicts that may be caused by the new process. In other words, it refers to whether the path can effectively avoid and resolve conflicts to ensure smooth delivery.

[0137] In detail, the conflict verification engine relies on server-side simulation and detection algorithms to load the process logic and resource configuration of intermediate optimization paths. Simultaneously, it extracts data such as potential interaction conflicts and conflict points identified in the early stages from the conflict database to construct a conflict verification scenario.

[0138] Specifically, the optimized execution path is a process plan that ultimately guides the actual execution of intellectual property delivery, clarifying the steps, resources, and conflict resolution methods. It is an "operation guide" that the system outputs to the execution end (such as reviewers and delivery teams).

[0139] In detail, intermediate optimized paths that have passed conflict compatibility verification are selected. These paths are then standardized: following the output template of the intellectual property delivery process, including step sequence, division of responsibilities, resource links, and conflict contingency plans, path details are supplemented and improved, such as assigning specific execution positions and related system operation entry points to each step. Then, through the system interface, the optimized execution path is output to the execution end, automatically triggering process tasks and ensuring that delivery is implemented according to the optimized path.

[0140] Conflict propagation module 104: used to aggregate task conflict traces of the delivery request based on the logical chain of the optimized execution path and the task priority, and to construct a conflict propagation topology graph of the delivery request through the task conflict traces.

[0141] In this embodiment, the logical chain based on the optimized execution path and the task priority to aggregate the task conflict traces of the delivery request includes:

[0142] Analyze the sequence of task nodes in the optimized execution path and calculate the sequence offset of the task nodes in the sequence;

[0143] When the sequence offset exceeds a preset conflict detection threshold, the corresponding task node is marked as a potential conflict trigger point;

[0144] By integrating the interaction chains of the potential conflict trigger points through the logical chain, the task conflict trace of the delivery request is obtained.

[0145] Specifically, the optimized execution path is a previously generated and verified intellectual property delivery process path, which contains an ordered sequence of task nodes and is a flowchart of the analysis.

[0146] The task node sequence is a sequence of task steps arranged in order in the optimized execution path, such as the review node and the correction node. Each node corresponds to an operation step in the intellectual property delivery and is the smallest execution unit of the process.

[0147] Sequence offset is the degree of deviation between the actual execution order and position of task nodes and the node sequence in the preset execution path. For example, if the preset node sequence is A→B→C, and the actual execution is A→C→B, the algorithm calculates the quantified value of the position and sequence deviation to reflect the degree of deviation in the process execution.

[0148] In detail, the conflict detection threshold is a preset critical value for sequence offset based on historical intellectual property delivery data and process stability requirements. When the offset exceeds this value, it indicates that the process execution deviates from the normal mode and may trigger a conflict. It serves as a baseline for judging conflict risk.

[0149] Potential conflict trigger points are task nodes marked as potentially initiating conflicts due to excessive sequence offsets. These nodes are the nascent points of conflict, and their impact on the process needs to be analyzed in detail later.

[0150] Specifically, the stored conflict detection threshold is obtained from historical conflict analysis results and process optimization configurations, and the sequence offset calculated in the previous step is compared with it. If the offset is greater than the threshold, the marking logic is triggered, and the corresponding node is labeled as a potential conflict trigger point in the task node data structure, while the process context at the time of triggering is recorded.

[0151] In detail, the logic chain is a chain of relationships between task nodes formed based on task priority and execution path logic, reflecting the business logic and conflict transmission rules of the intellectual property delivery process.

[0152] An interaction chain is the mutual influence between potential conflict triggers caused by process logic, resource dependencies, and other factors.

[0153] Task conflict traces are records of the trajectory of conflict generation and propagation in the process after integrating the interaction chain, including the conflict trigger point, transmission path, and scope of impact.

[0154] In this embodiment, constructing the conflict propagation topology map of the delivery request through the task conflict traces includes:

[0155] Analyze the conflict intensity of the task conflict traces;

[0156] Identify areas where the conflict intensity exceeds a preset propagation threshold, and mark the areas exceeding the threshold as conflict outbreak areas;

[0157] Connect the conflict outbreak areas and simultaneously extend the conflict range along the task dependency direction of the delivery request to obtain the initial diffusion path of the delivery request;

[0158] The integrity of the preliminary diffusion path is verified, and the verified preliminary diffusion paths are pieced together to form a conflict diffusion topology map.

[0159] Specifically, conflict intensity is a quantitative indicator that measures the extent to which conflict, as a trace of conflict within a task, affects the intellectual property delivery process. Influenced by factors such as the scale of resources involved in the conflict, the number of task nodes affected, and the risk of delivery delays, a higher value indicates greater destructive power of the conflict.

[0160] The preset propagation threshold is a critical value for conflict intensity based on historical experience in managing intellectual property delivery conflicts and process stability requirements. When the conflict intensity exceeds this value, it indicates that the conflict has reached the point of potential outbreak and requires close monitoring.

[0161] The conflict outbreak area is a process area consisting of one or more fragments of task conflict traces where the conflict intensity exceeds a threshold, and it is a concentrated outbreak of conflict.

[0162] In the intellectual property delivery process, the direction of the dependency relationship between task nodes due to resources and logic is the channel for conflict propagation.

[0163] The initial diffusion path is the path that connects the conflict outbreak area and extends along the mission-dependent direction, resulting in the possible propagation of the conflict. It includes the outbreak area and the extended scope of conflict impact, and is a candidate route for conflict diffusion.

[0164] Furthermore, integrity verification checks whether the initial diffusion path covers all potential conflict propagation areas and whether it conforms to task dependency logic and conflict propagation rules, ensuring that the path can fully present the overall picture of conflict propagation. It is a quality control step for the accuracy of the topology map.

[0165] The conflict propagation topology map is a visualized conflict propagation network formed by piecing together the initial propagation paths after verification. It includes conflict nodes, outbreak areas, extension nodes, propagation edges, and task-dependent directions, and is a map of the laws governing the propagation of intellectual property delivery conflicts.

[0166] Delivery module 105: used to resolve the task conflict traces based on the conflict diffusion topology graph, and deploy the optimal delivery path based on the resolution result.

[0167] In this embodiment, resolving the task conflict traces based on the conflict diffusion topology graph includes:

[0168] Analyze the distribution of conflict hotspots in the aforementioned conflict diffusion topology map;

[0169] Based on the task priority, the order of the diffusion paths of the conflict hotspots in the conflict hotspot distribution is adjusted, and the result of the adjustment is used as a preliminary resolution plan.

[0170] Verify the completeness of the task conflict traces in the preliminary resolution plan. When the completeness of the resolution exceeds a preset stability threshold, integrate and output the resolved task conflict traces.

[0171] Specifically, the distribution of conflict hotspots is the location, number, and interrelationship of nodes or regions with high conflict intensity and large impact range in the conflict diffusion topology map, reflecting the thermal distribution of concentrated conflict outbreaks.

[0172] The order of the spread path of conflict hotspots is determined by the order in which the conflict hotspots are formed in the topological graph according to the logic of conflict propagation, such as hotspot A → hotspot B → hotspot C, which determines the timing of the conflict spread.

[0173] The preliminary resolution plan is a preliminary conflict resolution scheme formed after adjusting the order of the conflict hotspots' spread paths. The task conflict trace is a record of the trajectory of the conflict's generation and spread in the intellectual property delivery process, and it is the object of resolution.

[0174] Furthermore, the completeness of resolution measures the degree to which the initial resolution plan addresses the traces of task conflicts. It is affected by factors such as the number of conflicts resolved and the degree of impact of residual conflicts. The higher the value, the more thorough the resolution.

[0175] The preset stability threshold is a critical value for the completeness of conflict resolution, set by the system based on the stability requirements of the intellectual property delivery process. When the completeness exceeds this value, it indicates that the conflict resolution effect has met the standard.

[0176] The resolved task conflict traces are trace records that have been processed by the initial resolution plan, have significantly reduced the impact of the conflict, and meet the stability threshold requirements. They serve as an acceptance report for the conflict resolution results.

[0177] After obtaining the traces of resolved task conflicts, the system algorithm combines the optimized execution path, task priority and other data to replan the delivery path: prioritize process nodes with thorough conflict resolution and high resource adaptability, and avoid areas affected by residual conflicts.

[0178] By using a path evaluation algorithm that considers delivery time, resource costs, and conflict risks, the optimal delivery path is selected and deployed to the execution end through a system interface to guide the actual operation of intellectual property delivery and ensure efficient and low-conflict execution.

[0179] In this embodiment, deploying the optimal delivery path based on the resolution result includes:

[0180] The resource nodes in the resolved task conflict traces are aggregated into key resource nodes:

[0181] Based on the task priority, the key resource nodes are connected, and a preliminary delivery path sequence is generated along the gradient guidance trajectory of the task priority;

[0182] Verify the execution efficiency integrity of the initial delivery path sequence. When the efficiency gain value meets the dynamic deployment threshold, reconstruct and output the optimal delivery path.

[0183] Specifically, the resolved task conflict traces are the trace records after the conflict has been resolved and the impact of the conflict has been significantly reduced.

[0184] Resource nodes are digital mappings of various resources involved in the intellectual property delivery process, such as server computing power, review personnel, and document storage units.

[0185] Key resource nodes are resource nodes that have a significant impact on the intellectual property delivery process and are closely related to task priority, such as dedicated review queues and high-computing-power servers that high-priority tasks rely on.

[0186] Gradient-guided trajectories are the gradient changes in task priority from high to low, reflecting the order of importance of task execution.

[0187] The initial delivery path sequence is a preliminary delivery process path formed by connecting key resource nodes according to task priority gradients, such as the execution order of high priority resources → medium priority resources → low priority resources.

[0188] Execution efficiency integrity is a comprehensive indicator that measures the efficiency of resource utilization and the timeliness of task completion when the initial delivery path sequence is executed. It is affected by factors such as the response speed of resource nodes and path redundancy. The higher the value, the more efficient the path.

[0189] Efficiency gain is a quantified measure of the efficiency improvement (such as time reduction rate and resource saving rate) of the initial delivery path sequence compared to historical / baseline paths, reflecting the effectiveness of path optimization.

[0190] The dynamic deployment threshold is a critical value for efficiency gain that is dynamically adjusted based on the efficiency requirements of the intellectual property delivery process. When the efficiency gain value exceeds this value, it indicates that the path execution efficiency meets the standard, and it is the standard value for judging whether the path is optimal.

[0191] The optimal delivery path is the delivery process path that, after verification, meets the efficiency gain requirements and has high execution efficiency and integrity. It is the final route for the system to be deployed to the execution end.

[0192] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0193] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0194] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0195] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0196] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0197] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An intelligent management system for intellectual property delivery based on an AI large-scale model, characterized in that: The system includes: Attribute acquisition module: used to receive intellectual property delivery requests and document resources, and synchronously acquire the key feature elements of the delivery request and the type attributes of the document resources; Temporary solution generation module: used to dynamically identify potential interaction conflicts between the key feature elements and the type attributes, and generate a temporary correction solution for the delivery request based on the potential interaction conflicts; Optimization module: used to dynamically allocate the task priority of the delivery request based on the temporary correction scheme, and simultaneously optimize the execution path of the delivery request based on the task priority; Conflict propagation module: used to aggregate task conflict traces of the delivery request based on the logical chain of the optimized execution path and the task priority, and to construct a conflict propagation topology graph of the delivery request through the task conflict traces; Delivery module: used to resolve the task conflict traces based on the conflict diffusion topology graph, and deploy the optimal delivery path based on the resolution result.

2. The intelligent management system for intellectual property delivery based on an AI large model as described in claim 1, characterized in that, The process of synchronously acquiring the key feature elements of the delivery request and the type attributes of the document resource includes: The content features of the delivery request are analyzed, the key feature elements of the delivery request are extracted from the content features, and the type attributes of the document resource are reviewed simultaneously.

3. The intelligent management system for intellectual property delivery based on an AI large model as described in claim 1, characterized in that, The dynamic identification of potential interaction conflicts between the key feature elements and the type attributes includes: Construct the rule conflict coefficient matrix of the delivery request based on the entropy difference results of the key feature elements and the type attributes; Extend the structured causal chain topology along the rule conflict coefficient matrix; The results of the topology extension are used to identify potential conflict propagation paths in the delivery request; Verify the pattern anomaly of the conflict propagation path, locate the endpoint of the abnormal propagation path of the delivery request based on the pattern anomaly, and identify the endpoint of the abnormal propagation path as a potential interaction conflict of the delivery request.

4. The intelligent management system for intellectual property delivery based on an AI large model as described in claim 1, characterized in that, The provisional correction scheme for generating the delivery request based on the potential interaction conflict includes: Based on the potential interaction conflicts, identify the conflict points between the key feature elements and the type attributes; By correcting the historical baseline data of the delivery request based on the conflict points, a preliminary correction proposal for the delivery request is generated; Based on the aforementioned preliminary revised proposal, the environmental parameters were subjected to feasibility simulation verification to obtain the conflict simulation screening results. The conflict simulation screening is integrated into a temporary correction scheme.

5. The intelligent management system for intellectual property delivery based on an AI large model as described in claim 4, characterized in that, The task priority of the delivery request is dynamically allocated based on the temporary correction scheme, wherein the calculation formula for dynamic allocation is: in: As a task priority, Let be the conflict entropy decay function. For the first The element deviation degree at each point of conflict For the first The path coupling coefficient at each conflict point As a dynamic correction factor, The maturity level of the temporary revised scheme is... For time-sensitive multipliers, and For learnable parameters, As the knowledge intensity decay factor, The total number of conflict points. For indexing.

6. The intelligent management system for intellectual property delivery based on an AI large model as described in claim 5, characterized in that, The process of simultaneously optimizing the execution path of the delivery request based on the task priority includes: Based on the task priority and the temporary correction scheme, an intermediate optimized path for the delivery request is generated; Verify the conflict compatibility of the intermediate optimization path; The intermediate optimization path that passes the verification is output as the optimized execution path.

7. The intelligent management system for intellectual property delivery based on an AI large model as described in claim 1, characterized in that, The logical chain that aggregates the task conflict traces of the delivery request based on the optimized execution path and the task priority includes: Analyze the sequence of task nodes in the optimized execution path and calculate the sequence offset of the task nodes in the sequence; When the sequence offset exceeds a preset conflict detection threshold, the corresponding task node is marked as a potential conflict trigger point; By integrating the interaction chains of the potential conflict trigger points through the logical chain, the task conflict trace of the delivery request is obtained.

8. The intelligent management system for intellectual property delivery based on an AI large model as described in claim 1, characterized in that, The step of constructing a conflict propagation topology map of the delivery request based on the task conflict traces includes: Analyze the conflict intensity of the task conflict traces; Identify areas where the conflict intensity exceeds a preset propagation threshold, and mark the areas exceeding the threshold as conflict outbreak areas; Connect the conflict outbreak areas and simultaneously extend the conflict range along the task dependency direction of the delivery request to obtain the initial diffusion path of the delivery request; The integrity of the preliminary diffusion path is verified, and the verified preliminary diffusion paths are pieced together to form a conflict diffusion topology map.

9. The intelligent management system for intellectual property delivery based on an AI large model as described in claim 1, characterized in that, The process of resolving the task conflict traces based on the conflict diffusion topology graph includes: Analyze the distribution of conflict hotspots in the aforementioned conflict diffusion topology map; Based on the task priority, the order of the diffusion paths of the conflict hotspots in the conflict hotspot distribution is adjusted, and the result of the adjustment is used as a preliminary resolution plan. Verify the completeness of the task conflict traces in the preliminary resolution plan. When the completeness of the resolution exceeds a preset stability threshold, integrate and output the resolved task conflict traces.

10. The intelligent management system for intellectual property delivery based on an AI large model as described in claim 1, characterized in that, The step of deploying the optimal delivery path based on the resolution results includes: The resource nodes in the resolved task conflict traces are aggregated into key resource nodes: Based on the task priority, the key resource nodes are connected, and a preliminary delivery path sequence is generated along the gradient guidance trajectory of the task priority; Verify the execution efficiency integrity of the initial delivery path sequence. When the efficiency gain value meets the dynamic deployment threshold, reconstruct and output the optimal delivery path.

Images