A scientific research project-based intellectual property management system

Abstract

The present application relates to the field of computer information technology, in particular to a kind of based on scientific research project's intellectual property management system, comprising: project intellectual property management module, for constructing the patentable knowledge graph of scientific research project, obtains series technical disclosure material;High-value intellectual property cultivation module, for establishing multidimensional evaluation mechanism, obtains patent layout score;Technology disclosure management module, for forming technology disclosure library, carries out patentable search and re-optimization integration, forms a series of patent proposal;If the patent proposal formed is batch proposal, the batch proposal is re-input into high-value intellectual property cultivation module, obtains new patent layout score;According to new patent layout score, sorting analysis is carried out, and whether high-value patent cultivation target can be achieved is re-evaluated.By the system, from technical achievements to patent proposal can be realized automatic mining and systematized layout, and the value of patent layout is quantitatively evaluated by multidimensional evaluation mechanism.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of computer information technology, in particular to a knowledge property right management system based on scientific research projects. BACKGROUND

[0002] At present, various innovative enterprises will apply for a large number of patents every year to protect their own innovation achievements. Therefore, the information management system for patent data has basically become a standard configuration of enterprises for managing massive patent data. The existing patent data management system mainly manages enterprise patent information, and basically manages the formed patents, such as visual management of patent assets, statutory period warning management of patents, cloud management of patent information, and interconnection management of patent data.

[0003] For example, in the prior art, the Chinese invention patent file with publication number CN113065844A and publication date July 2, 2021; the Chinese invention patent file with publication number CN116069760A and publication date May 5, 2023; the Chinese invention patent file with publication number CN117670597A and publication date March 8, 2024. The above patent files are related to the optimization of patent data management and related functions.

[0004] However, these systems mainly manage the formed patents of enterprises, and cannot fully cover the pre-stage knowledge property right planning, patent mining and layout, and cultivation of high-value knowledge property rights of scientific research projects. For the actual needs of innovative enterprises, how to plan, mine and layout the pre-stage knowledge property rights based on scientific research projects, and cultivate high-value knowledge property rights is particularly important.

[0005] In addition, in the pre-stage knowledge property right planning process, technical personnel usually write technical disclosure materials. This process has the following problems: many technical personnel do not understand the connotation and requirements of patents, and it is difficult to evaluate which technical achievements or technical data formed through scientific research projects can form patent points; technical personnel often write disclosure books for single-point technology when conceiving patents, which lacks systematicness and overall view, resulting in a lack of scientificity and systematicness in the formation of patents.

[0006] In addition, the patents formed based on the above technical disclosure books are mainly evaluated based on the patents themselves in the subsequent evaluation. However, for the transformation of achievements, a series of knowledge property rights composed of patents are often needed. Therefore, when carrying out knowledge property right layout around a scientific and technological achievement, the patent supervisor needs to analyze the value of the patent layout to determine how to achieve better transformation of achievements. However, there is no solution in the prior art that specifically addresses this need. SUMMARY

[0007] To solve the above technical problems, the present application provides a scientific research project-based intellectual property management system, which can realize the automatic mining and systematic layout from technical achievements to patent proposals starting from scientific research projects, and quantitatively evaluate the value of patent layout through a multi-dimensional evaluation mechanism, thereby improving the output quality and efficiency of intellectual property and supporting the efficient transformation of scientific and technological achievements.

[0008] The present application is realized by adopting the following technical solutions:

[0009] A scientific research project-based intellectual property management system comprises:

[0010] A project intellectual property management module is used to construct a patentable knowledge graph of scientific research projects and convert technical foundation data in each scientific research project into a series of technical disclosure materials.

[0011] A high-value intellectual property cultivation module is used to establish a multi-dimensional evaluation mechanism, conduct comprehensive evaluation of the patent layout of the series of technical disclosure materials of each scientific research project output by the project intellectual property management module, and obtain the patent layout score of each scientific research project.

[0012] A technical disclosure management module is used to form a technical disclosure library according to the series of technical disclosure materials of each scientific research project, and conduct patentable search and re-optimization and integration according to the corresponding patent layout score to form a series of patent proposals; the patent proposals include single proposals and batch proposals; if the formed patent proposal is a batch proposal, the batch proposal is re-input into the high-value intellectual property cultivation module for again conducting comprehensive evaluation of the patent layout, obtaining a new patent layout score, and conducting sorting analysis according to the new patent layout score to re-evaluate whether the high-value patent cultivation goal can be achieved.

[0013] Constructing a patentable knowledge graph of scientific research projects specifically comprises the following steps:

[0014] Step S1. Data collection and preprocessing;

[0015] Step S2. Constructing and training an entity-relation extraction model based on neural network;

[0016] Step S3. Using the trained entity-relation extraction model to extract multiple types of entities and the relationships between them, generating RDF triple data of entity-relation-entity, and storing them in a graph database to realize the construction of a data-based patentable knowledge graph.

[0017] The data preprocessing specifically refers to formatting processing and information extraction of data.

[0018] The entity-relation extraction model is based on a pre-training model BERT and combines position information of entities and relations.

[0019] The patent layout score is obtained by calculating the scores of six dimensions of the patent layout of each scientific research project, and taking the weighted average value as the patent layout score of each scientific research project.

[0020] The layout integrity specifically refers to the completeness of the patent layout, the layout effectiveness specifically refers to the efficiency of the patent layout of the scientific research project, the layout accuracy refers to the accuracy of hitting the technical key points of the patent layout, the layout uniqueness refers to the repetition of the technical key points of the patent layout of the scientific research project and other patent layouts or single-point patent technical key points, the layout consistency specifically refers to the matching of the technical key points of the patent layout of the scientific research project and the strategic direction of the company, and the layout timeliness specifically refers to whether the layout can be completed within a specified time limit based on the demand of the high-value patent cultivation target.

[0021] The score A of the layout integrity is obtained by querying the technical pedigree of the scientific research project to check whether the patent layout based on the scientific research project is complete.

[0022] ;

[0023] The score B of the layout effectiveness is obtained by:

[0024] B=(number of layout patents / average number of layout patents) 100%;

[0025] The score C of the layout accuracy is obtained by:

[0026] C=(number of hit technical key points / total number of technical key points) 100%;

[0027] The score D of the layout uniqueness is obtained by:

[0028] D=(number of repetitions / total number of comparisons) 100%;

[0029] The score E of the layout consistency is obtained by:

[0030] E=(number of matching technical key points and important branches / total number of technical key points) 100%;

[0031] The score F of the layout timeliness is obtained by:

[0032] .

[0033] The specific calculation method for the patent layout score V is as follows:

[0034] ,

[0035] In the formula, It is the first i Scores in each dimension It is the first i The weights of each dimension are adjustable.

[0036] Based on the corresponding patent layout score, patentability search and re-optimization and integration specifically refer to: sorting patent layouts according to their scores, determining the corresponding resource allocation and processing priority based on the scores; conducting patentability searches on the patent layouts according to the priorities, and re-optimizing and integrating the patent layouts based on the search results, patent application strategies, and technical levels.

[0037] The ranking analysis based on the new patent layout scores, and the reassessment of whether the high-value patent cultivation goal can be achieved, specifically refers to: ranking and analyzing the new patent layout scores, identifying the patent layouts that can achieve the high-value patent cultivation goal, and determining whether the patent layouts that do not achieve the high-value patent cultivation goal can be artificially or supplementarily explored to ensure that the high-value patent cultivation goal can be achieved.

[0038] It also includes a proposal management module and a patent management module; the proposal management module is used to receive single proposal applications output by the technology disclosure management module and batch proposal applications output by the high-value intellectual property cultivation module, and process the proposal applications; the patent management module is used to receive single or batch proposal filing applications output by the proposal management module, form patent cases, and manage patents.

[0039] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0040] 1. This invention first constructs a patentability knowledge graph of research projects, transforming a large amount of useful data generated in these projects into patentable technical disclosure materials. For researchers lacking patent awareness, this process can reduce the uncertainties and delays that may occur during manual technical disclosure mining. For researchers already skilled in technical disclosure mining, this process can reduce labor costs and make technical disclosure mining more scientific and systematic.

[0041] Based on this, the present invention further conducts a comprehensive evaluation of the patent layout of a series of technical disclosure materials, which helps patent officers identify combinations that may produce high-value patents. The scoring determines the corresponding resource allocation and processing priority, thereby optimizing resource allocation and improving overall work efficiency.

[0042] Further, since the technical personnel of a scientific research project manage different technical branches respectively, and the overall patent planning and layout often requires a certain overall situation and patent protection strategy, the core invention points or common invention points need to be condensed and extracted for multiple technical branches. Therefore, the present application forms a technology disclosure library, which can ensure that there is no omission in the technical invention, and then the patentability search and auxiliary layout are performed by an agent or a professional, according to the high-value patent cultivation target demand, a series of patent proposals are formed by re-splitting and combining, so as to avoid omission of technical points or repeated work.

[0043] Finally, the batch of proposals are re-input into the high-value intellectual property cultivation module, and the patent layout comprehensive evaluation is performed again, so as to re-evaluate whether the target demand can be achieved, and judge whether further precipitation or optimization is needed. If the target has been basically achieved, the intellectual property output and circulation can be more reasonable. In addition, through the high-value intellectual property cultivation module, the problem of strong randomness and high operation difficulty of traditional patent layout can be solved.

[0044] In summary, through the layer-by-layer progressive technical solutions, the present application not only simplifies the process steps of patent mining, but also provides a systematic layout for patent mining of scientific research projects of the same type. In addition, the application of the high-value intellectual property cultivation module further improves the scientificity and rationality of intellectual property management, and ensures the efficient output and circulation of intellectual property. This not only helps the successful transformation of scientific research projects, but also provides solid support for technical innovation in related fields.

[0045] 2、The entity-relation extraction model based on neural network used in the present application has good performance in relation extraction and joint extraction tasks. The model combines the advantages of pre-training language models (such as BERT) and graph neural networks (GNN), and can efficiently identify potential technical entities and their semantic relationships from unstructured text related to scientific research projects, and then generate structured knowledge representation, providing high-quality data support for the construction of subsequent patentable knowledge graph.

[0046] Further, the present application introduces a scientific research project patentable mining mechanism based on graph neural network and knowledge graph, which can solve the problems of uncertainty, incompleteness and high personnel requirements in traditional technology disclosure mining.

[0047] 3、In the present application, data preprocessing can not only quickly identify and extract technical data and other related auxiliary data that can be used to form patents in scientific research projects, but also provide a high-quality data foundation for subsequent entity-relation extraction model training.

[0048] 4, usually, the achievement transformation is not a single patent, but a series of patent-based intellectual property portfolio. Therefore, when arranging the intellectual property around a scientific and technological achievement, the patent administrator needs to evaluate and determine the reliability of the achievement transformation through the value degree analysis of the patent arrangement.

[0049] Therefore, the present application proposes a high-value intellectual property cultivation module based on a multi-dimensional evaluation mechanism, which aims to calculate the patent arrangement score of each scientific research project through a systematic method, and provide an important reference for the cultivation and output of high-value intellectual property.

[0050] Specifically, the module comprehensively evaluates the patent arrangement of each scientific research project from six different dimensions, which can ensure the comprehensive consideration of the value degree of the patent arrangement and avoid the deviation caused by a single indicator. At the same time, the module supports flexible adjustment of weights, allowing dynamic adjustment of the weights of each dimension according to the patent strategy needs of the company. Finally, based on the calculated patent arrangement score, the patent arrangement of the scientific research project is ranked, which can help quickly identify the intellectual property portfolio with high value potential.

[0051] 5, the present application also includes a proposal management module and a patent management module, which can solve the problem of insufficient system and comprehensiveness of technical proposal management before patent application. BRIEF DESCRIPTION OF DRAWINGS

[0052] The present application will be further described in detail below in conjunction with the drawings and specific embodiments of the present application, in which:

[0053] Figure 1 The present application is a structural schematic diagram. DETAILED DESCRIPTION

[0054] Embodiment 1

[0055] As a basic embodiment of the present application, the present application includes a scientific research project-based intellectual property management system, which includes a project intellectual property management module, a high-value intellectual property cultivation module and a technical disclosure management module.

[0056] The project intellectual property management module is used to construct the patentable knowledge graph of the scientific research project, and to convert the technical basic data in each scientific research project into a series of technical disclosure materials. The high-value intellectual property cultivation module is used to establish a multi-dimensional evaluation mechanism, to comprehensively evaluate the patent arrangement of the series of technical disclosure materials of each scientific research project output by the project intellectual property management module, and to obtain the patent arrangement score of each scientific research project. The technical disclosure management module is used to form a technical disclosure library according to the series of technical disclosure materials of each scientific research project, and to perform patentable search and re-optimization and integration according to the corresponding patent arrangement score, to form a series of patent proposals.

[0057] The patent proposal includes a single proposal and a batch proposal. If the formed patent proposal is a batch proposal, the batch proposal is re-input into the high-value intellectual property cultivation module, and the patent layout comprehensive evaluation is performed again to obtain a new patent layout score. And according to the new patent layout score, the sorting analysis is carried out, and it is re-evaluated whether the high-value patent cultivation target can be reached, and it is judged whether it needs to be further deposited or optimized.

[0058] Embodiment 2

[0059] As a preferred embodiment of the present application, the present application comprises a knowledge property right management system based on scientific research projects, comprising a project intellectual property management module, a high-value intellectual property cultivation module and a technology disclosure management module.

[0060] The project intellectual property management module is used to construct a patentable knowledge graph of scientific research projects, and to convert the technical basis data in each scientific research project into a series of technology disclosure materials. Among them, constructing the patentable knowledge graph of scientific research projects specifically includes the following steps:

[0061] Step S1. Data collection and preprocessing;

[0062] Step S2. Construct and train an entity-relation extraction model based on neural network;

[0063] Step S3. Use the trained entity-relation extraction model to extract multiple entities and their relationships, generate RDF triple data of entity-relation-entity, and store it in a graph database to realize the construction of a data-based patentable knowledge graph.

[0064] The high-value intellectual property cultivation module is used to establish a multi-dimensional evaluation mechanism, and to perform a patent layout comprehensive evaluation on the series of technology disclosure materials of each scientific research project output by the project intellectual property management module, to obtain a patent layout score of each scientific research project. Among them, the method for obtaining the patent layout score is: calculating the six-dimensional score of each scientific research project patent layout, and taking the weighted average value as the patent layout score of each scientific research project. The six dimensions are layout integrity, layout effectiveness, layout accuracy, target consistency, layout uniqueness and layout timeliness.

[0065] The technology disclosure management module is used to form a technology disclosure library according to the series of technology disclosure materials of each scientific research project, and to perform patentable search and re-optimization and integration according to the corresponding patent layout score, to form a series of patent proposals.

[0066] The patent proposal includes a single piece proposal and a batch proposal. If the formed patent proposal is a batch proposal, the batch proposal is re-input into the high-value intellectual property cultivation module, and the patent layout comprehensive evaluation is performed again to obtain a new patent layout score. And according to the new patent layout score, the sorting analysis is performed, and it is re-evaluated whether the high-value patent cultivation target can be reached.

[0067] Embodiment 3

[0068] As another preferred embodiment of the present application, the present application includes a knowledge property right management system based on scientific research projects, comprising a project intellectual property management module, a high-value intellectual property cultivation module, and a technology disclosure management module. The project intellectual property management module is used to construct a patentable knowledge graph of scientific research projects, and to convert the technical foundation data in each scientific research project into a series of technology disclosure materials. The high-value intellectual property cultivation module is used to establish a multi-dimensional evaluation mechanism, and to perform a patent layout comprehensive evaluation on the series of technology disclosure materials of each scientific research project output by the project intellectual property management module, to obtain a patent layout score of each scientific research project.

[0069] The technology disclosure management module is used to form a technology disclosure library according to the series of technology disclosure materials of each scientific research project, and to perform a patentable search and re-optimization integration according to the corresponding patent layout score, to form a series of patent proposals. Specifically, the patent layout is sorted according to the patent layout score, and the corresponding resource allocation and processing priority is determined according to the high and low of the patent layout score. The patent layout is searched for patentability according to the priority, and the patent layout is re-optimized and integrated based on the search results.

[0070] The patent proposal includes a single piece proposal and a batch proposal. If the formed patent proposal is a batch proposal, the batch proposal is re-input into the high-value intellectual property cultivation module, and the patent layout comprehensive evaluation is performed again to obtain a new patent layout score. According to the new patent layout score, the sorting analysis is performed, and it is re-evaluated whether the high-value patent cultivation target can be reached. That is, a complete and reliable patent layout is found, which can achieve the high-value patent cultivation target. It is judged whether the patent layout which is not complete and reliable can be obtained by manual or supplementary mining to ensure that the high-value patent cultivation target can be reached.

[0071] Embodiment 4

[0072] As another preferred embodiment of the present application, the present application includes a knowledge property right management system based on scientific research projects, referring to the description attached Figure 1 , comprising a project intellectual property management module, a high-value intellectual property cultivation module, a technology disclosure management module, a proposal management module, and a patent management module.

[0073] The project intellectual property management module is used to build a patentable knowledge graph of the scientific research project, and to convert technical basis data in each scientific research project into a series of technical disclosure materials. Specifically, it can include the following steps:

[0074] Step S1. Data collection and preprocessing.

[0075] The data collection specifically refers to collecting relevant technical basis data of the scientific research project. Taking the aircraft landing gear failure maintenance scientific research project as an example, it includes: ① fault data, including fault time, fault components, component symptoms, fault description, and fault causes; ② fault location, including shock absorber, load-bearing column, and retraction mechanism; ③ fault component symptoms, including vibration, oil leakage, fracture, and wear; ④ fault handling, including fault maintenance principle, fault maintenance process, fault maintenance method, and fault maintenance device.

[0076] By integrating the collected historical data, i.e. historical fault data, fault location, fault component symptoms, and fault handling process, as the data source for building the patentable knowledge graph of the scientific research project.

[0077] Before building the knowledge graph, the collected historical text data is preprocessed to find subsequent patentable technical data and other related auxiliary data. The data preprocessing specifically refers to formatting the data and information extraction.

[0078] Specifically, the data formatting specifically refers to converting files with inconsistent fonts and encodings into text files with unified encoding formats. Information extraction specifically refers to the construction of the patentable knowledge graph involving fault components, component symptoms, and fault handling. Therefore, the fault knowledge in the original historical files (excel, docx, doc, PDF) needs to be extracted to obtain an entity list, which serves as training data for the entity-relation extraction model.

[0079] Step S2. Construct and train a neural network-based entity-relation extraction model. The entity-relation extraction model is based on a pre-trained model BERT, and the entity and relation extraction model combined with the position information of entities and relations performs well for tasks such as relation extraction and joint extraction. Artificial entity-relation dataset construction is performed in advance, i.e., the original fault data is divided into multiple entities such as fault components, component symptoms, fault descriptions, and fault handling. The relationship can be represented as: fault component—(occurs)—component symptom, component symptom—(causes)—fault description, fault cause—(leads to)—fault description, and fault handling—(solves)—component symptom. Thus, the entity-relation-entity RDF (Resource Description Framework) triplets are obtained, which are used for training the entity-relation extraction model.

[0080] Step S3. Using the trained entity-relation extraction model, extract multiple entities and their relationships to generate entity-relation-entity RDF triplet data and store it in a graph database, thereby realizing the construction of a data-based patentability knowledge graph.

[0081] Specifically, in data preprocessing, the entity list containing fault parts, symptoms, and handling nodes is obtained through information extraction, which provides a vocabulary basis for constructing the knowledge graph nodes. Based on the self-training entity-relation extraction model, RDF triplet data of entity-relation-entity can be extracted from fault handling.

[0082] The constructed node and relationship knowledge are stored in the Neo4j graph database, which is a mature graph structure database. The Cypher declarative query language can be used to conveniently manage nodes and relationship edges. At the same time, Neo4j provides a friendly visualization tool that can facilitate intuitive inspection of the construction effect of the knowledge graph. The interface of Neo4j with deep learning frameworks such as TensorFlow will provide technical support for subsequent graph representation learning. Based on the Neo4j graph database, a data-based patentability knowledge graph is constructed.

[0083] In summary, the project intellectual property management module collects, preprocesses, trains, and constructs a knowledge graph based on a large amount of useful data generated in scientific research projects, thereby realizing the construction of a patentability knowledge graph and providing a convenient way for automatic mining of technical disclosure.

[0084] The high-value intellectual property cultivation module is used to establish a multi-dimensional evaluation mechanism, to comprehensively evaluate patent layout of a series of technical disclosure materials of each scientific research project output by the project intellectual property management module, to obtain a patent layout score of each scientific research project, and to facilitate a patent supervisor to identify a combination of possible high-value patents, to provide a reliable basis for batch layout and output of high-value intellectual property. For example, when intellectual property layout is carried out around a scientific and technological achievement, the patent supervisor can analyze and evaluate the reliability of the achievement transformation through the patent layout score.

[0085] Specifically, the method for obtaining the patent layout score is to calculate the score of six dimensions of the patent layout of each scientific research project, and to take the weighted average value as the patent layout score of each scientific research project. The six dimensions are layout integrity, layout effectiveness, layout accuracy, target consistency, layout uniqueness, and layout timeliness.

[0086] Further, the layout integrity specifically refers to the completeness of the patent layout. By querying the technical pedigree of the scientific research project, it is checked whether the patent layout based on the scientific research project is complete, reflecting the standard and completeness of the patent layout. Therefore, the score A of the layout integrity is:

[0087] .

[0088] The layout effectiveness specifically refers to the efficiency of the patent layout of the scientific research project. That is, when a scientific research project is laid out, the ratio of the total number of patent layout to the average number of patent layout is reflected to support the achievement transformation. For example, the average number of patent layout is 10. By comparing with the average number of patent layout, the effectiveness of the patent layout is reflected, and the value is a percentage. Therefore, the score B of the layout effectiveness is:

[0089] B=(the number of layout patents / the average number of layout patents) 100%。

[0090] The layout accuracy refers to the accuracy of the patent layout hitting the technical key points. Specifically, the patent supervisor compares the technical key words pre-prepared according to the scientific research project with the key words in the patent layout for the same or similar degree. The score C of the layout accuracy is a percentage, and the calculation method is:

[0091] C=(the number of hit technical key points / the total number of technical key points) 100%。

[0092] The layout uniqueness refers to the repetition of the technical key points of the patent layout of the scientific research project and other patent layouts or single-point patent technical key points, reflecting the uniqueness of the layout patent. The score D of the layout uniqueness is a percentage, and the calculation method is:

[0093] D = (number of repetitions / total number of comparisons) 100%.

[0094] Layout consistency specifically refers to the matching of the technical key points of the patent layout of the scientific research project with the strategic direction of the company. The process needs to confirm whether the technical key points of the patent layout come from the key branches of the company's technology map. If it comes from the key branches, then according to the matching degree with the key branches, the consistency is defined. The score E of layout consistency is a percentage, and the calculation method is:

[0095] E = (number of technical key points matched with important branches / total number of technical key points) 100%.

[0096] Layout timeliness specifically refers to the timeliness of the patent layout determined by the technical person in charge of the scientific research project based on the demand of high-value patent cultivation target, such as the application of achievement transformation project, standard essential patent cultivation, etc. The time limit requirement for completing the layout is determined, such as completing the layout within the time limit, then it is 1; if it is not completed within the time limit, then it is 0. Specifically, the score F of layout timeliness is:

[0097] .

[0098] Finally, the weighted average value is taken as the patent layout score V of each scientific research project:

[0099] ,

[0100] In the formula, is the score of the i th dimension, is the weight of the i th dimension, which can be adjusted according to the patent application target of the year.

[0101] The technology disclosure management module is used to form a technology disclosure library according to the series of technology disclosure materials of each scientific research project. Since the technical personnel of a scientific research project each manages different technical branches, and the overall patent planning and layout often requires a certain overall situation and patent protection strategy, the core invention points or common invention points need to be condensed and extracted from multiple technical branches, so forming a technology disclosure library can ensure that there is no omission of technical invention.

[0102] The technology disclosure management module is further configured to perform patentability search and re-optimization and integration according to the corresponding patent layout score to form a series of patent proposals. Specifically, the patent layout can be sorted according to the patent layout score, and the corresponding resource allocation and processing priority can be determined by the patent supervisor according to the patent layout score. The patent agency or professional personnel can perform patentability search on the patent layout according to the priority. According to the prior art searched by the patent agency or professional personnel, the search result can be obtained, that is, the technical points with creativity and the technical points disclosed by the prior art can be distinguished. Based on the search result, the patent application strategy and the technical level, a series of patent proposals are formed after further optimizing and integrating the technology disclosure materials.

[0103] The patent proposal includes a single proposal and a batch proposal. If the formed patent proposal is a single proposal, the proposal process is triggered to output a single proposal application. If the formed patent proposal is a batch proposal, the layout process is triggered, and the batch proposal is input into the high-value intellectual property cultivation module again to perform patent layout comprehensive evaluation to obtain a new patent layout score. The new patent layout score is sorted and analyzed to re-evaluate whether the high-value patent cultivation target can be achieved. Specifically, a complete and reliable patent layout that can achieve the high-value patent cultivation target can be quickly found. The patent layout that cannot achieve the high-value patent cultivation target is an incomplete and unreliable patent layout, and further judgment is made on whether the incomplete and unreliable patent layout can be obtained through manual or supplementary mining, or through re-batch layout to ensure that the high-value patent cultivation target can be achieved, so that a series of patents can be formed to support the achievement transformation project or the standard essential patent cultivation. After the batch layout is completed, the proposal process is triggered to output a batch proposal application.

[0104] The proposal management module is configured to receive the single proposal application output by the technology disclosure management module and the batch proposal application output by the high-value intellectual property cultivation module, process the proposal application, and have a proposal query function.

[0105] The patent management module is configured to receive the single proposal or batch proposal filing application output by the proposal management module, form a patent case, and perform subsequent patent management.

[0106] In summary, various corresponding transformation schemes made by those skilled in the art without creative mental effort according to the technical solutions and technical concepts of the present application after reading the present application file all belong to the scope protected by the present application.

Claims

1. A system for intellectual property management based on scientific research projects, characterized in that: The application comprises: a project intellectual property management module for constructing a patentable knowledge graph of a scientific research project and converting technical foundation data in each scientific research project into a series of technical disclosure materials; a high-value intellectual property cultivation module for establishing a multi-dimensional evaluation mechanism, performing comprehensive evaluation on the series of technical disclosure materials of each scientific research project output by the project intellectual property management module, and obtaining a patent layout score of each scientific research project; a technical disclosure management module for forming a technical disclosure library according to the series of technical disclosure materials of each scientific research project, and performing patentable search and re-optimization and integration according to the corresponding patent layout score to form a series of patent proposals; the patent proposals include single proposals and batch proposals; if the formed patent proposal is a batch proposal, the batch proposal is re-input into the high-value intellectual property cultivation module to perform comprehensive evaluation on the patent layout again to obtain a new patent layout score; and the new patent layout score is sorted and analyzed to re-evaluate whether the high-value patent cultivation target can be achieved; constructing a patentable knowledge graph of a scientific research project specifically comprises the following steps: Step S1. Data collection and preprocessing; Step S2. Constructing and training an entity-relation extraction model based on a neural network; Step S3. Extracting multiple entities and the relationships therebetween by using the trained entity-relation extraction model to generate RDF triple data of entity-relation-entity, and storing the data in a graph database to realize construction of a data-based patentable knowledge graph. 2.The system based on scientific research project intellectual property management according to claim 1, characterized in that: The data preprocessing specifically refers to format processing and information extraction of the data. 3.The system based on scientific research project intellectual property management according to claim 1, characterized in that: The entity-relation extraction model is based on a pre-trained BERT model combined with position information of entities and relationships. 4.The system based on scientific research project intellectual property management according to claim 1, characterized in that: The method for obtaining the patent layout score is to calculate the scores of six dimensions of the patent layout of each scientific research project, and take the weighted average value as the patent layout score of each scientific research project; the six dimensions are layout completeness, layout effectiveness, layout accuracy, target consistency, layout uniqueness, and layout timeliness.

5. The system according to claim 4, wherein: The layout completeness specifically refers to the completeness of the patent layout, the layout effectiveness specifically refers to the efficiency of the patent layout of the scientific research project, the layout accuracy refers to the accuracy of the patent layout hitting the technical key points, the layout uniqueness refers to the repetition of the technical key points of the patent layout of the scientific research project with other patent layouts or single-point patent technical key points, the layout consistency specifically refers to the matching of the technical key points of the patent layout of the scientific research project with the company strategic direction, and the layout timeliness specifically refers to whether the layout can be completed within a specified time limit based on the demand of the high-value patent cultivation target. 6.The system based on scientific research project intellectual property management according to claim 5, characterized in that: The score A of the layout completeness is obtained by querying the technical pedigree of the scientific research project to check whether the patent layout based on the scientific research project is complete. ； The score B of the layout effectiveness is obtained by checking the efficiency of the patent layout of the scientific research project. B = (number of layout patents / average number of layout patents) 100%; The score C of the layout accuracy is obtained by checking the accuracy of the patent layout hitting the technical key points. C = (number of hit technical keypoints / total number of technical keypoints) 100%; The score D of the layout uniqueness is obtained by checking the repetition of the technical key points of the patent layout of the scientific research project with other patent layouts or single-point patent technical key points. D = (number of repeats / total number of comparisons) 100%; The score E of the layout consistency is obtained by checking the matching of the technical key points of the patent layout of the scientific research project with the company strategic direction. E = (number of technology key points matched with important branches / total number of technology key points) 100%; The score F of the layout timeliness is obtained by checking whether the layout can be completed within a specified time limit based on the demand of the high-value patent cultivation target. 。 7.The system of claim 6, wherein: The specific calculation method of the patent layout score V is: ， wherein is the score for the i dimension, is the weight for the i dimension, which can be adjusted. 8.The system based on scientific research project intellectual property management according to claim 1, characterized in that: The patentability search and re-optimization and integration of the specific patent layout according to the corresponding patent layout score are as follows: the patent layout is sorted according to the patent layout score, the corresponding resource allocation and processing priority are determined according to the high and low of the patent layout score, the patentability search is performed on the patent layout according to the priority, and the patent layout is re-optimized and integrated based on the search result, the patent application strategy and the technical level. 9.The system based on scientific research project intellectual property management according to claim 1, characterized in that: The re-evaluation of whether the high-value patent cultivation target can be achieved according to the sorting analysis of the new patent layout score is as follows: the patent layout that can achieve the high-value patent cultivation target is found according to the sorting analysis of the new patent layout score, and it is judged whether the patent layout that cannot achieve the high-value patent cultivation target can achieve the high-value patent cultivation target through manual or supplementary mining, so as to ensure that the high-value patent cultivation target can be achieved.

10. The scientific research project-based intellectual property management system according to any one of claims 1-9, characterized in that: The application further comprises a proposal management module and a patent management module; the proposal management module is used for receiving the single proposal application output by the technology disclosure management module and the batch proposal application output by the high-value intellectual property cultivation module, processing the proposal application; the patent management module is used for receiving the single proposal or batch proposal case filing application output by the proposal management module, forming a patent case, and performing patent management.

Images