Process management method, system and equipment for enterprise patent application and approval and medium

Abstract

The invention discloses an enterprise patent application and approval process management method, system and device and a medium, and relates to the technical field of enterprise intellectual property management, and the method comprises the steps: obtaining a patent application document submitted by enterprise personnel, and intercepting a feature technology part in the patent application document through an OCR recognition technology; acquiring a technical keyword from the feature technology part by using a named entity recognition technology, and performing document retrieval processing on the patent application document according to the technical keyword; for the patent application document subjected to document retrieval processing, performing calculation by using a preset SHAP value interpretation model to obtain a decision contribution of the characteristics of the technical keywords in the patent application document; according to the decision contribution of each technical keyword in the patent application document, obtaining an approval result of the patent application document; the problems of low processing efficiency, low retrieval accuracy, high approval error rate and the like of current manual management are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise intellectual property management, and more specifically, to a process management method, system, equipment and medium for enterprise patent application and approval. Background Art

[0002] The enterprise patent management department is responsible for managing the patent work of each subsidiary. The interface departments of each subsidiary will submit the relevant patent application project information, and the enterprise intellectual property management department will do the preliminary screening. The current patent application and approval have the following problems: 1. The traditional patent application process relies on manual review, which is inefficient and prone to errors; 2. Patent searches rely on keyword matching, which makes it difficult to identify similarities in technical solutions; 3. The approval process lacks intelligent decision-making support, making it difficult to quantify and evaluate the value of patents; 4. The problem of data silos in multi-department collaboration is serious, and status tracking is difficult. Summary of the invention

[0003] The purpose of the present invention is to provide a process management method, system, equipment and medium for enterprise patent application and approval to solve the problems existing in the above-mentioned background technology.

[0004] The above technical objectives of the present invention are achieved through the following technical solutions: First, this application provides a process management method for enterprise patent application and approval, including the following specific steps: Obtain patent application documents submitted by enterprise personnel, and extract the characteristic technical parts in the patent application documents through OCR recognition technology; Use named entity recognition technology to obtain technical keywords from the characteristic technical part, and perform document retrieval processing on the patent application documents based on the technical keywords; For patent application documents processed through document retrieval, the decision contribution of the features of technical keywords in the patent application documents is calculated using the preset SHAP value interpretation model; According to the decision-making contribution of each technical keyword in the patent application document, the approval result of the patent application document is obtained, and the approval result includes approval or failure.

[0005] Based on the above technical solution, the present invention can also be improved as follows.

[0006] Furthermore, the above-mentioned document retrieval process of the patent application document based on technical keywords is specifically as follows: The weighted Jaccard algorithm is used to calculate the first similarity between each technical keyword in the patent application document and the keywords in other public documents; Generate semantic embedding vectors of patent application documents and other public documents using a preset BERT model, and calculate the second similarity between the patent application documents and other public documents; The comprehensive similarity of the patent application document is calculated by the first similarity and the second similarity. If the comprehensive similarity of the patent application document does not exceed a threshold, the patent application document is processed through document retrieval.

[0007] Furthermore, the first similarity is specifically: ; In the formula, A collection of technical keywords representing patent application documents A Keyword collection with other public documents B The first similarity of Indicates keywords i The weight of Indicates keywords j The weight of The second similarity is specifically: ; In the formula, Indicates patent application documents With other public documents The second similarity is Indicates patent application documents The semantic embedding vector of Indicates other public documents The semantic embedding vector of .

[0008] Furthermore, the above comprehensive similarity is specifically: ; In the formula, Indicates patent application documents With other public documents The comprehensive similarity of represents adjustable parameters, A collection of technical keywords representing patent application documents A Keyword collection with other public documents B The first similarity of Indicates patent application documents With other public documents The second similarity.

[0009] Furthermore, the decision-making contribution of the features of the technical keywords in the above patent application documents is specifically as follows: ; In the formula, Characteristics of technical keywords Contribution to decision making, F represents the set of all features, S represents a feature subset, is the SHAP value explaining the model on the feature subset S The predicted value of Represents the feature set in the feature vector composed of technical keywords.

[0010] Furthermore, the above approval results are as follows: ; In the formula, Indicates the approval result. Characteristics of technical keywords Contribution to decision making, is the bias term, represents the number of features, represents a symbolic function that maps weighted sums to binary decisions.

[0011] Furthermore, the method further includes: obtaining the information gain of the features of each technical keyword and the patent life cycle value of the patent application document based on a preset dynamic decision tree model, wherein: The information gain is specifically: ; In the formula, Indicates patent application documents Characteristics of technical keywords The information gain of Indicates patent application documents The information entropy of It is a characteristic of technical keywords The value is A subset of It is a characteristic of technical keywords All possible values ​​of ; The specific value of the patent life cycle is: ; In the formula, represents the lifetime value of a patent, Indicates that the patent is t Expected revenue per year, d represents the discount rate, T Represents the life cycle of a patent.

[0012] In a second aspect, the present application provides a process management system for enterprise patent application and approval, which is applied to the process management method for enterprise patent application and approval in any one of the first aspects, including: Intelligent application module, used to obtain patent application documents submitted by enterprise personnel, and intercept the characteristic technical part of patent application documents through OCR recognition technology; Patent search module, used to obtain technical keywords from the characteristic technical part by using named entity recognition technology, and perform document search processing on patent application documents according to the technical keywords; A data analysis module is used to calculate the decision contribution of the features of technical keywords in the patent application documents processed through document retrieval using a preset SHAP value interpretation model; The process approval module is used to obtain the approval results of the patent application document based on the decision contribution of each technical keyword in the patent application document. The approval results include approval or failure.

[0013] In a third aspect, the present application provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any one of the methods in the first aspect when executing the computer program.

[0014] In a fourth aspect, the present application provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions enable a computer to execute any one of the methods in the first aspect.

[0015] Compared with the prior art, the present invention has at least the following beneficial effects: In this application, the process management method and system provided solve the problems of cumbersome work process in query and statistics, repeated checking between staff when aggregating relevant data, and information asymmetry in daily management that inevitably leads to omissions due to the scattered distribution of data in business and management departments compared to manual management methods. This solution allows business personnel and management personnel to browse and query on one management platform, and when problems arise, they can be modified and processed in a timely manner, greatly improving work efficiency.

[0016] In this application, duplication of patent project applications can also be avoided. Since the enterprise patent management department is responsible for managing the patent work of each subsidiary, the interface department of each subsidiary will submit the relevant patent application project information in a summary, and the enterprise intellectual property management department will do preliminary screening. Because there are multiple subordinate units in the same industry and business, repeated patent applications are inevitable. Through this plan, patent management personnel at all levels can inquire about all patent applications of other units at all levels. When submitting relevant patent information, staff of competent departments at all levels can make preliminary selections, and can also learn from the patent technologies of other units to optimize and solve related technical problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings: Figure 1 A method flow chart of a management method in an embodiment of the present invention; Figure 2 This is a connection diagram of a management system in an embodiment of the present invention; Figure 3 1 is a functional module diagram of a management system in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0019] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0021] In the description of the embodiments of the present invention, "plurality" means at least 2.

[0022] Embodiment 1: In order to solve the problems of low processing efficiency, low search accuracy, high approval error rate, etc. in current manual management, this embodiment provides a process management method for enterprise patent application and approval, such as Figure 1 As shown, the specific steps include: S1, obtain the patent application documents submitted by enterprise personnel, and use OCR recognition technology to capture the characteristic technical part of the patent application documents.

[0023] S2, using named entity recognition technology to obtain technical keywords from the characteristic technical part, and perform document retrieval processing on the patent application document based on the technical keywords.

[0024] Optionally, the above-mentioned document retrieval process of the patent application document based on the technical keywords is specifically as follows: S21, using a weighted Jaccard algorithm to calculate the first similarity between each technical keyword in the patent application document and the keywords in other public documents.

[0025] The first similarity is specifically: ; In the formula, A collection of technical keywords representing patent application documents A Keyword collection with other public documents B The first similarity of Indicates keywords i The weight of Indicates keywords j The weight of .

[0026] S22, using a preset BERT model to generate semantic embedding vectors of the patent application document and other public documents, and calculating the second similarity between the patent application document and other public documents.

[0027] The second similarity is specifically: ; In the formula, Indicates patent application documents With other public documents The second similarity is Indicates patent application documents The semantic embedding vector of Indicates other public documents The semantic embedding vector of .

[0028] S23, calculating the comprehensive similarity of the patent application document by using the first similarity and the second similarity. If the comprehensive similarity of the patent application document does not exceed a threshold, the patent application document is processed through document retrieval.

[0029] The above comprehensive similarity is specifically: ; In the formula, Indicates patent application documents With other public documents The comprehensive similarity of represents adjustable parameters, A collection of technical keywords representing patent application documents A Keyword collection with other public documents B The first similarity of Indicates patent application documents With other public documents The second similarity.

[0030] S3, for the patent application documents processed by document retrieval, the decision contribution of the features of the technical keywords in the patent application documents is calculated using the preset SHAP value interpretation model.

[0031] Among them, the decision-making contribution of the features of the technical keywords in the above patent application documents is specifically: ; In the formula, Characteristics of technical keywords Contribution to decision making, F represents the set of all features, S represents a feature subset, is the SHAP value explaining the model on the feature subset S The predicted value of Represents the feature set in the feature vector composed of technical keywords.

[0032] S4, according to the decision contribution of each technical keyword in the patent application document, the approval result of the patent application document is obtained, and the approval result includes approval or failure.

[0033] The above approval results are as follows: ; In the formula, Indicates the approval result. Characteristics of technical keywords Contribution to decision making, is the bias term, represents the number of features, represents a symbolic function that maps weighted sums to binary decisions.

[0034] Optionally, the method further includes: obtaining the information gain of the features of each technical keyword and the patent life cycle value of the patent application document based on a preset dynamic decision tree model, wherein: The information gain is specifically: ; In the formula, Indicates patent application documents Characteristics of technical keywords The information gain of Indicates patent application documents The information entropy of It is a characteristic of technical keywords The value is A subset of It is a characteristic of technical keywords All possible values ​​of .

[0035] The above patent life cycle value is specifically: ; In the formula, represents the lifetime value of a patent, Indicates that the patent is t Expected annual revenue, d represents the discount rate, T Represents the life cycle of a patent.

[0036] Embodiment 2: This application provides a process management system for enterprise patent application and approval, which is applied to the process management method for enterprise patent application and approval in Embodiment 1, such as Figure 2 As shown, including: Intelligent application module, used to obtain patent application documents submitted by enterprise personnel, and intercept the characteristic technical part of patent application documents through OCR recognition technology; The patent search module is used to obtain technical keywords from the characteristic technical part by using named entity recognition technology, and perform document search processing on the patent application document according to the technical keywords; the patent search module may include: The first subunit is used to calculate the first similarity between each technical keyword in the patent application document and the keywords in other public documents by using the weighted Jaccard algorithm; the above first similarity is specifically: ; In the formula, A collection of technical keywords representing patent application documents A Keyword collection with other public documents B The first similarity of Indicates keywords i The weight of Indicates keywords j The weight of .

[0037] The second subunit is used to generate semantic embedding vectors of patent application documents and other public documents using a preset BERT model, and calculate the second similarity between the patent application documents and the other public documents; the second similarity is specifically: ; In the formula, Indicates patent application documents With other public documents The second similarity is Indicates patent application documents The semantic embedding vector of Indicates other public documents The semantic embedding vector of .

[0038] The third subunit is used to calculate the comprehensive similarity of the patent application document by using the first similarity and the second similarity. If the comprehensive similarity of the patent application document does not exceed the threshold, the patent application document is processed by document retrieval. The above comprehensive similarity is specifically: ; In the formula, Indicates patent application documents With other public documents The comprehensive similarity of represents adjustable parameters, A collection of technical keywords representing patent application documents A Keyword collection with other public documents B The first similarity of Indicates patent application documents With other public documents The second similarity.

[0039] The data analysis module is used to calculate the decision contribution of the features of the technical keywords in the patent application document processed by the document retrieval using the preset SHAP value interpretation model; the decision contribution of the features of the technical keywords in the patent application document is specifically: ; In the formula, Characteristics of technical keywords Contribution to decision making, F represents the set of all features, S represents a feature subset, is the SHAP value explaining the model on the feature subset S The predicted value of Represents the feature set in the feature vector composed of technical keywords.

[0040] The process approval module is used to obtain the approval results of the patent application document according to the decision contribution of each technical keyword in the patent application document. The approval results include approval or failure. The above approval results are specifically: ; In the formula, Indicates the approval result. Characteristics of technical keywords Contribution to decision making, is the bias term, represents the number of features, represents a symbolic function that maps weighted sums to binary decisions.

[0041] Among them, the above-mentioned process management system can be divided into user layer, application layer and service layer in actual use. At the same time, the functions in the user layer, application layer and service layer are respectively implemented through one or more of the intelligent application module, intelligent application module, patent search module and process approval module.

[0042] For the user layer, such as Figure 3 As shown, it may include: 1.1. Inventor; Function: Inventors can submit patent applications through the system and fill in patent-related technical information, claims, etc. At the same time, inventors can view the approval status of patents in real time, understand the current approval progress, approval opinions, and whether additional materials are required.

[0043] Interaction mode: Provides a friendly user interface that supports functions such as file uploading, form filling, and status query.

[0044] 1.2. Reviewer; Function: The reviewer is responsible for reviewing the submitted patent application. The system will automatically generate approval suggestions based on the patent content. The reviewer can make approval decisions based on the system suggestions or his own judgment. The reviewer can also view the historical approval records of the patent, similar patents in related technical fields and other information to assist in decision-making.

[0045] Interaction mode: Provides functions such as approval task list, approval suggestions, and history query.

[0046] 1.3. Administrator; Function: The administrator is responsible for the overall management of the system, including user rights allocation, role management, system configuration, operation status monitoring, etc. The administrator can view the system usage, approval efficiency, error log, etc. to ensure the stable operation of the system.

[0047] Interaction mode: Provides functions such as user management, permission configuration, and system monitoring.

[0048] For the application layer, such as Figure 3 As shown, it may include: 2.1. User management module; Function: Manage the user roles and permissions of the system, perform multi-role permission control, and ensure that different roles (inventors, reviewers, administrators) can access the functional modules related to their responsibilities. Support user registration, login, permission allocation, role switching and other operations.

[0049] Technical implementation: Based on the RBAC (role-based access control) model, refined permission control is achieved through role-permission mapping.

[0050] Data flow: user login -> permission verification -> function access.

[0051] 2.2. Intelligent application module; Function: Provides structured patent application forms and supports automatic extraction of technical features in patent documents. After users upload documents, OCR extraction technology is used to identify the text content in patent documents, and named entity recognition (NER extraction) technology is used to extract key technical terms, inventor information, claims, etc. to generate application forms.

[0052] Technical implementation: OCR recognition technology (such as Tesseract), NER model (such as BERT-based NER).

[0053] Data flow: User uploads document -> OCR extracts text -> NER extracts keywords -> generates application form.

[0054] 2.3. Patent search module; Function: Supports semantic search and similarity matching. Users can enter queries (keywords or upload patent documents) to search for related patents. The system will return a list of patents with the highest similarity and provide detailed information on the patents (such as title, abstract, claims, etc.).

[0055] Technical implementation: The weighted Jaccard algorithm calculates text similarity, the BERT model generates semantic embedding vectors, and the Neo4j graph database stores patent technology associations.

[0056] Data flow: user input query -> feature extraction -> similarity calculation -> return result.

[0057] For the service layer, such as Figure 3 As shown, it may include: 3.1. Intelligent search service; Function: Provide patent semantic search and similarity calculation services, and support users to search for relevant patents by keywords or patent documents.

[0058] Technical implementation: The weighted Jaccard algorithm calculates text similarity, and the BERT model generates semantic embedding vectors.

[0059] Data flow: receive query request -> calculate similarity -> return search results.

[0060] 3.2. Approval decision service; Function: Provide automated approval decision support and generate approval recommendations based on patent characteristics.

[0061] Technical implementation: dynamic decision tree model + SHAP value explanation model; the dynamic decision tree model process is application submission and feature analysis, and the SHAP value explanation model process is decision tree approval and output results.

[0062] Data flow: Receive approval request -> feature analysis -> generate approval suggestion.

[0063] 3.3. Status tracking service; Function: Provide real-time update of patent status and historical record query services to ensure the traceability of the approval process.

[0064] Technical implementation: blockchain evidence storage technology.

[0065] Data flow: state update -> blockchain record -> real-time synchronization.

[0066] 3.4. Data analysis service; Function: Provide patent value assessment and examination cycle prediction services to support corporate decision-making.

[0067] Technical implementation: LTV algorithm + ARIMA time series prediction model.

[0068] Data flow: data input -> model calculation -> output analysis results.

[0069] Specifically, through the above-mentioned user layer, application layer and service layer, the problems of cumbersome work process in query and statistics, repeated checking between staff when summarizing related data, and the inevitable omissions due to information asymmetry in daily management due to the scattered distribution of data in business and management departments are solved; this system can enable business personnel and management personnel to browse and query on one management platform, and when problems arise, they can be modified and handled in time, greatly improving work efficiency.

[0070] Furthermore, it can also avoid duplication of patent project applications. Since the enterprise patent management department is responsible for managing the patent work of each subsidiary, the interface department of each subsidiary will submit the relevant patent application project information in a summary, and the enterprise intellectual property management department will do preliminary screening. Because there are multiple subordinate units in the same industry and business, it is inevitable that there will be repeated patent applications. Through this system, the patent management personnel of units at all levels can inquire about all the patent applications of other units at all levels. When submitting relevant patent information, the staff of the competent departments at all levels can make preliminary selections, and they can also learn from the patent technologies of other units to optimize and solve related technical problems.

[0071] Embodiment 3: The embodiment of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the method of Embodiment 1 is implemented when the processor executes the computer program.

[0072] Embodiment 4: The embodiment of the present application provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions enable a computer to execute the method of Embodiment 1.

[0073] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The process management method for enterprise patent application and approval is characterized by: The specific steps include: Obtain patent application documents submitted by enterprise personnel, and extract the characteristic technical parts in the patent application documents through OCR recognition technology; Using named entity recognition technology to obtain technical keywords from the characteristic technical part, and performing document retrieval processing on the patent application document according to the technical keywords; For patent application documents processed through document retrieval, the decision contribution of the features of technical keywords in the patent application documents is calculated using the preset SHAP value interpretation model; According to the decision contribution of each technical keyword in the patent application document, the approval result of the patent application document is obtained, and the approval result includes approval passed and approval failed.

2. The process management method for enterprise patent application and approval according to claim 1 is characterized in that: The patent application documents are searched and processed according to the technical keywords, specifically: The weighted Jaccard algorithm is used to calculate the first similarity between each technical keyword in the patent application document and the keywords in other public documents; Generate semantic embedding vectors of patent application documents and other public documents using a preset BERT model, and calculate the second similarity between the patent application documents and other public documents; The comprehensive similarity of the patent application document is calculated by using the first similarity and the second similarity. If the comprehensive similarity of the patent application document does not exceed a threshold, the patent application document is processed through document retrieval.

3. The process management method for enterprise patent application and approval according to claim 2 is characterized in that: The first similarity is specifically: ； In the formula, A collection of technical keywords representing patent application documents A Keyword collection with other public documents B The first similarity of Indicates keywords i The weight of Indicates keywords j The weight of The second similarity is specifically: ； In the formula, Indicates patent application documents With other public documents The second similarity is Indicates patent application documents The semantic embedding vector of Indicates other public documents The semantic embedding vector of .

4. The process management method for enterprise patent application and approval according to claim 3 is characterized in that: The comprehensive similarity is specifically: ； In the formula, Indicates patent application documents With other public documents The comprehensive similarity of represents adjustable parameters, A collection of technical keywords representing patent application documents A Keyword collection with other public documents B The first similarity of Indicates patent application documents With other public documents The second similarity.

5. The process management method for enterprise patent application and approval according to claim 1 is characterized in that: The decision-making contribution of the features of the technical keywords in the patent application document is specifically: ； In the formula, Characteristics of technical keywords Contribution to decision making, F represents the set of all features, S represents a feature subset, is the SHAP value explaining the model on the feature subset S The predicted value of Represents the feature set in the feature vector composed of technical keywords.

6. The process management method for enterprise patent application and approval according to claim 1 is characterized in that: The approval results are as follows: ； In the formula, Indicates the approval result. Characteristics of technical keywords Contribution to decision making, is the bias term, represents the number of features, represents a symbolic function that maps weighted sums to binary decisions.

7. The process management method for enterprise patent application and approval according to claim 1 is characterized in that: The method further includes: obtaining the information gain of the features of each technical keyword and the patent life cycle value of the patent application document based on a preset dynamic decision tree model, wherein: The information gain is specifically: ； In the formula, Indicates patent application documents Characteristics of technical keywords The information gain of Indicates patent application documents The information entropy of It is a characteristic of technical keywords The value is A subset of It is a characteristic of technical keywords All possible values ​​of ; The patent life cycle value is specifically: ； In the formula, represents the lifetime value of a patent, Indicates that the patent is t Expected annual revenue, d represents the discount rate, T Represents the life cycle of a patent.

8. A process management system for enterprise patent application and approval, applied to a process management method for enterprise patent application and approval as claimed in any one of claims 1 to 7, characterized in that: include: Intelligent application module, used to obtain patent application documents submitted by enterprise personnel, and intercept the characteristic technical part of patent application documents through OCR recognition technology; A patent search module, used to obtain technical keywords from the characteristic technical part by using named entity recognition technology, and perform document search processing on the patent application document according to the technical keywords; A data analysis module is used to calculate the decision contribution of the features of technical keywords in the patent application documents processed through document retrieval using a preset SHAP value interpretation model; The process approval module is used to obtain the approval result of the patent application document according to the decision contribution of each technical keyword in the patent application document, and the approval result includes approval passed and approval failed.

9. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the method according to any one of claims 1 to 7 is implemented when the processor executes the computer program.

10. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, which enable a computer to execute the method of any one of claims 1-7.

Images