A knowledge service system for intellectual property protection

Abstract

The present invention relates to the field of knowledge service systems, and specifically to a knowledge service system for intellectual property protection. The present invention extracts basic data of plaintiffs and defendants by setting up a case analysis module, classifies first intellectual property data in the basic data, and obtains second intellectual property data by word segmentation extraction, and finally calculates the first overlap degree, the second overlap degree and the average infringement probability respectively, and gives corresponding results for users to refer to, so that users can obtain corresponding reference opinions when facing case disputes involving intellectual property rights, so as to facilitate users to better handle infringement disputes related to intellectual property rights; and solves the problem that the existing legal knowledge service system has a relatively large error when predicting the judgment result of the current case due to the difficulty in determining the validity of intellectual property rights in the case trial process.

Description

Technical Field

[0001] The present invention relates to the field of knowledge service systems, and in particular to a knowledge service system oriented to intellectual property protection. Background Art

[0002] At present, there are a variety of legal knowledge service systems that can infer the possible judgment results of current cases based on the legal documents of previous cases. However, for cases involving intellectual property rights, especially patent disputes, the validity of intellectual property rights such as patents often has a decisive reference value for the judgment results of the case. However, during the trial of the case, the parties can change the validity of the intellectual property rights through relevant legal channels, such as invalidating the patent. However, the legal documents often do not contain the detailed content of the relevant intellectual property rights. Therefore, the existing legal knowledge service system is difficult to determine the validity of intellectual property rights during the trial of the case, and thus the error is relatively large when predicting the judgment results of the current case. Therefore, the result of whether to continue the appeal is often quite different from the actual situation. Summary of the invention

[0003] In view of the deficiencies of the prior art, the present invention provides a knowledge service system for intellectual property protection, which solves the problem that the existing legal knowledge service system has difficulty in determining the validity of intellectual property rights during the case trial process, resulting in relatively large errors when predicting the judgment results of the current case.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A knowledge service system for intellectual property protection, the system comprising:

[0006] A data retrieval module, the data retrieval module is used to obtain retrieval content corresponding to the terms according to a plurality of preset terms and corresponding retrieval methods and paths;

[0007] A knowledge graph module, which is used to obtain legal documents from the court and patent information of the enterprise and to construct and update the knowledge graph;

[0008] The case analysis module is used to calculate the average infringement probability of the defendant based on the basic data containing the plaintiff and defendant information input by the user, and output a prompt message that the defendant has infringed when the average infringement probability is greater than the infringement threshold, and output the relevant calculation basis, otherwise, output a prompt message that there is no infringement.

[0009] Preferably, the data retrieval module comprises:

[0010] A legal search module, which is used to obtain corresponding legal provisions based on keywords input by users;

[0011] A case search module, which is used to obtain a number of cases according to the keywords input by the user, and sort the cases according to the number of keywords contained in each case;

[0012] A comprehensive search module is used to obtain corresponding search content according to the search formula input by the user.

[0013] Preferably, the case analysis module includes:

[0014] A plaintiff information input module, wherein the plaintiff information input module is used for a user to input the plaintiff's identity information, and obtains the corresponding first intellectual property data and related case information from the knowledge graph module according to the plaintiff's identity information, and performs word segmentation extraction on the first intellectual property data to obtain the second intellectual property data;

[0015] A defendant information input module, wherein the defendant information input module is used for a user to input the defendant's identity information and corresponding infringing product data;

[0016] The infringement analysis module is used to calculate the second overlap between the second intellectual property data and the infringing product data input by the plaintiff information input module, and establish a corresponding relationship between the second overlap and the average infringement probability to calculate the average infringement probability of the defendant, and output a prompt message of the defendant's infringement when the average infringement probability is greater than the infringement threshold, and output the corresponding infringement basis, otherwise, output a prompt message of non-infringement.

[0017] Preferably, the plaintiff information input module includes:

[0018] A plaintiff information input module, which is used to input the plaintiff's identity information;

[0019] A case extraction module, which is used to obtain the plaintiff's historical case data based on the knowledge graph;

[0020] The intellectual property extraction module is used to obtain the corresponding first intellectual property data from the knowledge graph according to the plaintiff's identity information, and perform word segmentation extraction on the first intellectual property data to obtain the second intellectual property data.

[0021] Preferably, the intellectual property extraction module comprises:

[0022] A classification module, the classification module is used to obtain the corresponding first intellectual property data from the knowledge graph module according to the plaintiff identity information input by the user, and classify the first intellectual property data into two categories according to whether the first intellectual property data is a patent;

[0023] A first extraction module, the first extraction module is used to obtain the first intellectual property data classified as a patent by the classification module, and obtain the claims of the patent document corresponding to the first intellectual property data, and the reference relationship between the claims, extract the technical solutions of each claim in the claims in turn, and construct a claim tree of the reference relationship between the claims;

[0024] The second extraction module is used to perform word segmentation extraction on several groups of technical solutions in the first extraction module and the first intellectual property data that are not classified as patents in the classification module through the multi-head attention mechanism of the Transformer model to obtain second intellectual property data.

[0025] Preferably, the infringement analysis module includes:

[0026] A vectorized encoding module, wherein the vectorized encoding module is used to vectorize the second intellectual property data and the third intellectual property data;

[0027] A first overlap module, the first overlap module is used to calculate the first overlap of the patent, and screen out the second intellectual property data classified as patents whose overlap with the existing patents is less than the overlap threshold, so as to generate third intellectual property data;

[0028] A second coincidence module, the second coincidence module is used to calculate the second coincidence between the third intellectual property data and the infringing product data, and calculate the second coincidence between the second intellectual property data not classified as a patent part and the infringing product data;

[0029] An infringement probability corresponding module, the infringement probability corresponding module is used to calculate the average infringement probability of the defendant according to the second coincidence degree of historical cases and the number of wins and losses of corresponding cases;

[0030] A result output module is used to output the average infringement probability, and when the average infringement probability is greater than the infringement threshold, output the prompt information of the defendant's infringement, and output the relevant calculation basis; otherwise, output the prompt information of non-infringement and the data source for the corresponding second overlap calculation.

[0031] Preferably, the first overlap module comprises:

[0032] An information transfer amount calculation module is used to calculate the information transfer amount of each technical solution in the first extraction module according to the number of claims involved; the calculation formula of the information transfer amount is:

[0033] N(i)=α(i)×β(i)

[0034] In the above formula, N(i) represents the information transmission amount of the i-th technical solution, α(i) represents the number of claims contained in the i-th technical solution, and β(i) represents the minimum value of the longest path length from one vertex to other vertices in the claim tree of the i-th technical solution;

[0035] An information importance calculation module is used to calculate the information importance of any technical solution according to the claim tree; the calculation formula of the information importance is:

[0036]

[0037] In the above formula, Z(i) represents the information importance of the i-th technical solution, D(V m , V n ) means containing word V m and V n The edge weights that appear in the claim tree, F j Indicates that it also contains the word V m and V n The number of times the j-th edge of appears in the claim tree. There are K different edges in the claim tree. S represents the claim tree.

[0038] A first overlap calculation module, the first overlap calculation module is used to calculate the first overlap of two patents according to the output results of the information importance calculation module and the information transmission amount calculation module;

[0039] A screening module is used to select second intellectual property data classified as patents whose first degree of overlap with existing patents is less than a threshold of overlap, so as to generate third intellectual property data.

[0040] Preferably, the calculation formula of the second overlap degree is:

[0041]

[0042] In the above formula, D represents the second coincidence degree, cos(x,y) represents the cosine value of two feature vectors x and y, the two feature vectors x and y correspond to the feature vectors of each technical solution in the two comparison objects, n represents the dimension of the feature vector, x i and i Represent the components of the two eigenvectors x and y respectively.

[0043] Preferably, the infringement probability corresponding module includes:

[0044] A historical data extraction module, which is used to extract legal documents of historical intellectual property infringement cases, and calculate the second overlap of each historical intellectual property infringement case through the vectorized coding module, the first overlap module and the second overlap module, and count the number of times the plaintiff won and lost the case according to the judgment results of the legal documents;

[0045] The infringement probability calibration module is used to set a plurality of overlap ranges of the second overlap degree, and calculate the average infringement probability of the second overlap degree within the overlap range according to the number of times the plaintiff wins and loses the case; the calculation formula of the average infringement probability is:

[0046]

[0047] In the above formula, P σ represents the average infringement probability of the second overlap degree of the σth overlap range, and They respectively represent the number of times the plaintiffs won and lost the historical cases corresponding to all the second overlap degrees within the σth overlap range.

[0048] Preferably, the calculation formula of the first overlap degree is:

[0049]

[0050] in,

[0051]

[0052] In the above formula, Indicates two patents Q 1 and Q 2 Similarity of claims, and Represent the vectors of the claims of two different patents, represents the vector of any patent claim, S represents the claim tree, Z(i) represents the information importance of the i-th technical solution, N(i) represents the information transmission amount of the i-th technical solution, A vector representing the i-th technical solution.

[0053] Compared with the prior art, the present invention provides a knowledge service system for intellectual property protection, which has the following beneficial effects:

[0054] 1. The present invention sets up a case analysis module to extract the basic data of the plaintiff and the defendant, classifies the first intellectual property data in the basic data, and obtains the second intellectual property data through word segmentation extraction, and finally calculates the first overlap degree, the second overlap degree and the average infringement probability respectively, and gives corresponding results for users to refer to, so that users can obtain corresponding reference opinions when facing case disputes involving intellectual property rights, so as to facilitate users to better handle infringement disputes related to intellectual property rights.

[0055] 2. The present invention sets up an infringement analysis module, uses a vectorized encoding module to perform vectorized conversion on the second intellectual property data and the third intellectual property data, calculates the first overlap of the plaintiff's patent through a first overlap module, and screens out the third intellectual property data, and finally calculates the second overlap between the third intellectual property data and the infringing product data, as well as the second overlap between the second intellectual property data that is not classified as the patent part and the infringing product data, so as to facilitate the subsequent establishment of a corresponding relationship with the defendant's average infringement probability.

[0056] 3. The present invention sets up a module corresponding to the infringement probability, so that users can intuitively derive the plaintiff's probability of winning based on previous cases, and output the average infringement probability, as well as the data source for the corresponding second coincidence calculation, so that users can conduct further checks and save time for searching information. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0058] Figure 1 It is a structural block diagram of the knowledge service system of the present invention;

[0059] Figure 2 It is a structural block diagram of the plaintiff information input module of the present invention;

[0060] Figure 3 It is a structural block diagram of the infringement analysis module of the present invention;

[0061] Figure 4 This is a schematic diagram of the results of word segmentation extraction;

[0062] Figure 5 Schematic diagram of the knowledge graph, plaintiff information and defendant information.

[0063] In the figure: 1. Data retrieval module; 11. Law and regulation retrieval module; 12. Case retrieval module; 13. Comprehensive retrieval module; 2. Knowledge graph module; 3. Case analysis module; 31. Plaintiff information input module; 311. Plaintiff information entry module; 312. Case extraction module; 313. Intellectual property extraction module; 3131. Classification module; 3132. First extraction module; 3133. Second extraction module; 32. Defendant information input module; 33. Infringement analysis module; 331. Vectorized coding module; 332. First overlap module; 3321. Information transmission calculation module; 3322. Information importance calculation module; 3323. First overlap calculation module; 3324. Screening module; 333. Second overlap module; 334. Infringement probability corresponding module; 3341. Historical data extraction module; 3342. Infringement probability calibration module; 335. Result output module. DETAILED DESCRIPTION

[0064] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods, so that the implementation process of how the present application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0065] Those skilled in the art can understand that all or part of the steps in the following embodiments can be completed by instructing the relevant hardware through a program, so the present application can be in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application can be in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0066] In order to solve the problem that the existing legal knowledge service system has a relatively large error when predicting the judgment result of the current case due to the difficulty in determining the validity of intellectual property rights in the case trial process, the present invention provides a knowledge service system for intellectual property protection, which predicts the validity of intellectual property rights in the trial process, thereby achieving the purpose of assisting in predicting the judgment result, so as to better provide a reference for the intellectual property protection of enterprises. Please refer to Figure 1-Figure 3 , the system comprises:

[0067] A data retrieval module 1 is used to obtain search content corresponding to a term according to a plurality of preset terms and corresponding search methods and paths; the data retrieval module 1 obtains relevant legal knowledge through keywords input by a user, and specifically includes the following modules: a regulation retrieval module 11 is used to obtain corresponding legal and regulatory provisions according to the keywords input by a user; a case retrieval module 12 is used to obtain a plurality of cases according to the keywords input by a user, and sort the cases according to the number of keywords contained in each case; a comprehensive retrieval module 13 is used to obtain corresponding search content according to the search formula input by a user.

[0068] A knowledge graph module 2 for obtaining the court's legal documents and the enterprise's patent information and constructing and updating the knowledge graph; a case analysis module 3 for calculating the average infringement probability of the defendant based on the basic data containing the plaintiff and defendant information input by the user, and outputting a prompt message of the defendant's infringement when the average infringement probability is greater than the infringement threshold, and outputting the relevant calculation basis, otherwise, outputting a prompt message of non-infringement, so as to facilitate manual further judgment based on the calculation basis. The case analysis module 3 needs to extract and pre-process the basic data first, and realize the purpose of predicting the average infringement probability of the defendant through calculation, so the case analysis module 3 further includes:

[0069] The plaintiff information input module 31 is used for the user to input the plaintiff's identity information, and obtains the corresponding first intellectual property data and related case information from the knowledge graph module 2 according to the plaintiff's identity information, and performs word segmentation extraction on the first intellectual property data to obtain the second intellectual property data. For example, NLPIR is used to perform word segmentation processing on the first intellectual property data through Chinese word segmentation, part-of-speech tagging, and named entity recognition, and keywords in the first intellectual property data are extracted. The extracted keywords are as follows: Figure 4 As shown; the plaintiff information input module 31 needs to complete the automatic extraction and preprocessing of the plaintiff information to facilitate the subsequent further calculation to obtain the second intellectual property data. The plaintiff information input module 31 further includes the following modules:

[0070] A plaintiff information input module 311 for inputting the plaintiff's identity information; a case extraction module 312 for obtaining the plaintiff's historical case data from a knowledge graph or a related network path according to the plaintiff's information input by the plaintiff information input module 311, such as Figure 5As shown, the plaintiff information and defendant information and the knowledge graph are obtained; the intellectual property extraction module 313 is used to obtain the corresponding first intellectual property data from the knowledge graph according to the plaintiff identity information, and extract the first intellectual property data by word segmentation to obtain the second intellectual property data. The intellectual property extraction module 313 needs to distinguish whether the first intellectual property data is a patent, so as to better classify and extract different intellectual property rights, so as to facilitate the subsequent more targeted calculation of the second overlap degree. Therefore, the intellectual property extraction module 313 includes:

[0071] A classification module 3131 is used to obtain the corresponding first intellectual property data from the knowledge graph module 2 according to the plaintiff identity information input by the user, and classify the first intellectual property data into two categories according to whether the first intellectual property data is a patent; a first extraction module 3132 is used to obtain the first intellectual property data classified as a patent by the classification module 3131, and obtain the claims of the patent document corresponding to the first intellectual property data, and the citation relationship of the claims, extract the technical solutions of each claim in the claims in turn, and construct a claim tree of the citation relationship of each claim; a second extraction module 3133 is used to perform word segmentation extraction on several groups of technical solutions in the first extraction module 3132 and the first intellectual property data that is not classified as a patent in the classification module 3131 through the multi-head attention mechanism of the Transformer model to obtain the second intellectual property data.

[0072] A defendant information input module 32 for a user to input the identity information of the defendant and the corresponding infringing product data, wherein the infringing product data includes the defendant's product data, such as a mechanical structure or a process method, etc.; an infringement analysis module 33 for calculating the second overlap between the second intellectual property data and the infringing product data input by the plaintiff information input module 31, and establishing a corresponding relationship between the second overlap and the average infringement probability to calculate the average infringement probability of the defendant, and outputting a prompt message of the defendant's infringement when the average infringement probability is greater than the infringement threshold, and outputting a corresponding infringement basis, otherwise, outputting a prompt message of non-infringement. The infringement analysis module 33 needs to process the data extracted by the plaintiff information input module 31 and the defendant information input module 32 to calculate the average infringement probability of the defendant, so the infringement analysis module 33 includes:

[0073] A vectorized encoding module 331 for vectorized encoding of the second intellectual property data and the third intellectual property data; a first overlap module 332 for calculating a first overlap degree of patents and screening out the second intellectual property data classified as patents whose overlap degree with the existing patents is less than an overlap degree threshold, so as to generate the third intellectual property data;

[0074] The first overlap module 332 further processes the intellectual property data related to the patent to calculate the first overlap. The first overlap module 332 includes: an information transfer amount calculation module 3321 for calculating the information transfer amount according to the number of claims involved in each technical solution in the first extraction module 3132; the calculation formula of the information transfer amount is:

[0075] N(i)=α(i)×β(i)

[0076] In the above formula, N(i) represents the information transmission amount of the i-th technical solution, α(i) represents the number of claims contained in the i-th technical solution, and β(i) represents the minimum value of the longest path length from one vertex to other vertices in the claim tree of the i-th technical solution;

[0077] An information importance calculation module 3322 is used to calculate the information importance of any technical solution according to the claim tree; the calculation formula of the information importance is:

[0078]

[0079] In the above formula, Z(i) represents the information importance of the i-th technical solution, D(V m , V n ) means containing word V m and V n The edge weights that appear in the claim tree, F j Indicates that it also contains the word V m and V n The number of times the j-th edge of appears in the claim tree. There are K different edges in the claim tree. S represents the claim tree.

[0080] The first overlap calculation module 3323 is used to calculate the first overlap of two patents according to the output results of the information importance calculation module 3322 and the information transmission amount calculation module 3321; the calculation formula of the first overlap is:

[0081]

[0082] in,

[0083]

[0084] In the above formula, Indicates two patents Q 1 and Q 2 Similarity of claims, and Represent the vectors of the claims of two different patents, represents the vector of any patent claim, S represents the claim tree, Z(i) represents the information importance of the i-th technical solution, N(i) represents the information transmission amount of the i-th technical solution, A vector representing the i-th technical solution.

[0085] A screening module 3324 is used to select second intellectual property data classified as patents whose first degree of overlap with existing patents is less than an overlap threshold to generate third intellectual property data.

[0086] A second overlap module 333 is used to calculate the second overlap between the third intellectual property data and the infringing product data, and to calculate the second overlap between the second intellectual property data not classified as the patent part and the infringing product data; the calculation formula of the second overlap is:

[0087]

[0088] In the above formula, D represents the second coincidence degree, cos(x,y) represents the cosine value of two feature vectors x and y, the two feature vectors x and y correspond to the feature vectors of each technical solution in the two comparison objects, n represents the dimension of the feature vector, x i and i Represent the components of the two eigenvectors x and y respectively.

[0089] The infringement probability corresponding module 334 is used to calculate the average infringement probability of the defendant according to the second overlap degree of historical cases and the number of wins and losses of the corresponding cases; the infringement probability corresponding module 334 needs to collect statistics on historical cases to achieve the correspondence between the second overlap degree and the average infringement probability, so the average infringement probability corresponding module 334 includes:

[0090] A historical data extraction module 3341 for extracting legal documents of historical intellectual property infringement cases, and calculating the second overlap of each historical intellectual property infringement case through the vectorized coding module 331, the first overlap module 332 and the second overlap module 333, and counting the number of times the plaintiff won and lost the case according to the judgment results of the legal documents;

[0091] The infringement probability calibration module 3342 is used to set the overlap range of several second overlap degrees, and calculate the average infringement probability of the second overlap degrees within the overlap range according to the number of times the plaintiff wins and loses; the calculation formula of the average infringement probability is:

[0092]

[0093] In the above formula, P σ represents the average infringement probability of the second overlap degree of the σth overlap range, and They respectively represent the number of times the plaintiffs won and lost the historical cases corresponding to all the second overlap degrees within the σth overlap range.

[0094] The result output module 335 is used to output the average infringement probability, and when the average infringement probability is greater than the infringement threshold, output the prompt information of the defendant's infringement, and output the relevant calculation basis. Otherwise, it outputs the prompt information of non-infringement, and the data source for the corresponding second overlap calculation. The infringement threshold is generally set to 50%.

[0095] The above implementation methods have been described in detail. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A knowledge service system for intellectual property protection, characterized in that: The system includes: A data retrieval module (1), the data retrieval module (1) being used to obtain retrieval content corresponding to the terms according to a plurality of preset terms and corresponding retrieval methods and paths; A knowledge graph module (2), wherein the knowledge graph module (2) is used to obtain legal documents from the court and patent information of the enterprise and to construct and update the knowledge graph; A case analysis module (3), the case analysis module (3) is used to calculate the average infringement probability of the defendant based on the basic data containing the plaintiff and defendant information input by the user, and output a prompt message indicating that the defendant has infringed the rights of the defendant when the average infringement probability is greater than an infringement threshold, and output a relevant calculation basis; otherwise, output a prompt message indicating that the defendant has not infringed the rights of the defendant; The case analysis module (3) includes: A plaintiff information input module (31), wherein the plaintiff information input module (31) is used for a user to input the plaintiff's identity information, and to obtain corresponding first intellectual property data and related case information from the knowledge graph module (2) based on the plaintiff's identity information, and to perform word segmentation extraction on the first intellectual property data to obtain second intellectual property data; A defendant information input module (32), wherein the defendant information input module (32) is used by a user to input the identity information of the defendant and corresponding infringing product data; An infringement analysis module (33), the infringement analysis module (33) being used to calculate the second overlap between the second intellectual property data and the infringing product data input by the plaintiff information input module (31), and to establish a corresponding relationship between the second overlap and the average infringement probability, so as to calculate the average infringement probability of the defendant, and to output a prompt message indicating that the defendant has infringed when the average infringement probability is greater than an infringement threshold, and to output a corresponding infringement basis, otherwise, to output a prompt message indicating that the defendant has infringed; The infringement analysis module (33) comprises: A vectorized encoding module (331), the vectorized encoding module (331) being used to perform vectorized encoding on the second intellectual property data and the third intellectual property data; The first overlap module (332) is used to calculate the first overlap of the patent, and screen out the second intellectual property data classified as patents whose overlap with the existing patent is less than the overlap threshold, so as to generate the third intellectual property data; the calculation formula of the first overlap is: ； in, ； In the above formula, Indicates two patents and Similarity of claims, and Represent the vectors of the claims of two different patents, A vector representing any patent claim, represents the claim tree, The information importance of the i-th technical solution, represents the amount of information transferred by the i-th technical solution, A vector representing the i-th technical solution; The second overlap module (333) is used to calculate the second overlap between the third intellectual property data and the infringing product data, and to calculate the second overlap between the second intellectual property data not classified as the patent part and the infringing product data; the calculation formula of the second overlap is: ； In the above formula, D represents the second degree of coincidence, Represents two eigenvectors and The cosine value of and The feature vectors corresponding to each technical solution in the two comparison objects, represents the dimension of the feature vector, and Represents two eigenvectors respectively and The amount of An infringement probability corresponding module (334), the infringement probability corresponding module (334) is used to calculate the average infringement probability of the defendant according to the second coincidence degree of historical cases and the number of wins and losses of corresponding cases; A result output module (335), the result output module (335) is used to output an average infringement probability, and when the average infringement probability is greater than an infringement threshold, output prompt information indicating that the defendant has infringed, and output relevant calculation basis; otherwise, output prompt information indicating that there is no infringement, and a data source for calculating the corresponding second coincidence degree; The infringement probability corresponding module (334) includes: A historical data extraction module (3341), the historical data extraction module (3341) is used to extract legal documents of historical intellectual property infringement cases, and calculate the second overlap of each historical intellectual property infringement case through the vectorized coding module (331), the first overlap module (332) and the second overlap module (333), and count the number of times the plaintiff won and lost the case according to the judgment results of the legal documents; The infringement probability calibration module (3342) is used to set a plurality of overlap ranges of the second overlap degree, and calculate the average infringement probability of the second overlap degree within the overlap range according to the number of times the plaintiff wins and loses the case; the calculation formula of the average infringement probability is: ； In the above formula, Indicates The average infringement probability of the second overlap degree of the overlap range is and Respectively represent the The number of times the plaintiff won and lost the historical cases corresponding to all the second degrees of overlap within the overlap range.

2. The knowledge service system according to claim 1, characterized in that: The data retrieval module (1) comprises: A legal retrieval module (11), the legal retrieval module (11) is used to obtain corresponding legal provisions according to keywords input by a user; A case search module (12), the case search module (12) being used to obtain a number of cases according to keywords input by a user, and to sort the cases according to the number of keywords contained in each case; A comprehensive search module (13), wherein the comprehensive search module (13) is used to obtain corresponding search content according to a search formula input by a user.

3. The knowledge service system according to claim 1, characterized in that: The plaintiff information input module (31) comprises: A plaintiff information input module (311), wherein the plaintiff information input module (311) is used to input the plaintiff's identity information; A case extraction module (312), wherein the case extraction module (312) is used to obtain the plaintiff's historical case data based on the knowledge graph; The intellectual property extraction module (313) is used to obtain the corresponding first intellectual property data from the knowledge graph according to the plaintiff's identity information, and to perform word segmentation extraction on the first intellectual property data to obtain the second intellectual property data.

4. The knowledge service system according to claim 3, characterized in that: The intellectual property extraction module (313) includes: A classification module (3131), the classification module (3131) is used to obtain the corresponding first intellectual property data from the knowledge graph module (2) according to the plaintiff identity information input by the user, and classify the first intellectual property data into two categories according to whether the first intellectual property data is a patent; A first extraction module (3132), the first extraction module (3132) is used to obtain the first intellectual property data classified as a patent by the classification module (3131), and obtain the claims of the patent document corresponding to the first intellectual property data, and the reference relationship between the claims, extract the technical solutions of each claim in the claims in turn, and construct a claim tree of the reference relationship between each claim; The second extraction module (3133) is used to perform word segmentation extraction on several groups of technical solutions in the first extraction module (3132) and the first intellectual property data that are not classified as patents in the classification module (3131) through the multi-head attention mechanism of the Transformer model to obtain second intellectual property data.

5. The knowledge service system according to claim 1, characterized in that: The first coincidence module (332) comprises: An information transfer amount calculation module (3321), the information transfer amount calculation module (3321) is used to calculate the information transfer amount according to the number of claims involved in each technical solution in the first extraction module (3132); the calculation formula of the information transfer amount is: ； In the above formula, represents the amount of information transferred by the i-th technical solution, represents the number of claims included in the i-th technical solution, Represents the minimum value of the longest path length from one vertex to other vertices in the claim tree of the i-th technical solution; An information importance calculation module (3322), the information importance calculation module (3322) is used to calculate the information importance of any technical solution according to the claim tree; the calculation formula of the information importance is: ； In the above formula, represents the information importance of the i-th technical solution, Indicates that it contains words and The edge weights that appear in the claim tree, Indicates that it also contains the word and The number of times the j-th edge of appears in the claim tree. There are K different edges in the claim tree. Represents a claim tree; A first overlap calculation module (3323), the first overlap calculation module (3323) is used to calculate the first overlap of two patents according to the output results of the information importance calculation module (3322) and the information transmission amount calculation module (3321); A screening module (3324), wherein the screening module (3324) is used to select second intellectual property data classified as patents whose first degree of overlap with existing patents is less than a threshold of overlap, so as to generate third intellectual property data.

Images