A legal processing system based on big data and artificial intelligence

Abstract

The present invention relates to the technical field of legal affairs processing, and specifically discloses a legal affairs processing system based on big data and artificial intelligence, including: a registration module for obtaining the case type of each uploaded case; an extraction module for obtaining all extracted phrases of each uploaded case based on the case type of each uploaded case; a calculation module for obtaining all similar phrases of each uploaded case based on each extracted phrase of each uploaded case; a construction module for obtaining the case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on all similar phrases of each uploaded case; an analysis module for obtaining the legal affairs processing result of each uploaded case based on the case similarity value between each uploaded case and all similar cases of the corresponding uploaded case. The present invention realizes providing the uploader with a legal analysis text that is helpful for analyzing the uploaded case, thereby improving the efficiency of legal services.

Description

Technical Field

[0001] The present invention relates to the field of legal affairs processing technology, and in particular to a legal affairs processing system based on big data and artificial intelligence. Background Art

[0002] At present, with the development of artificial intelligence technology, it has become a trend to apply AI technology to the field of legal services. The traditional legal processing process is that lawyers conduct comprehensive review and communication to determine the legal processing opinions of the case. This method is not only inefficient, but also difficult to ensure the comprehensiveness and accuracy of the lawyers' legal processing opinions, resulting in low efficiency of the law firm's legal processing. The traditional legal processing process is no longer compatible with the increasing number of legal cases today. Therefore, how to introduce AI technology into the field of legal services to improve the efficiency and quality of legal services, and how to help lawyers give legal processing suggestions during the case handling process have become urgent issues.

[0003] However, the existing legal processing system based on big data and artificial intelligence only conducts corresponding assessments on the content of legal work and determines the performance of legal management through multiple factor indicators, but does not consider how to provide the uploader with all the legal analysis texts that are helpful for analyzing the uploaded case, and does not consider how to improve the efficiency of legal services. For example, the patent with the publication number "CN110414872A" and the patent name "Intelligent Legal Management Module, System and Method" includes the following steps: a central processing module and at least one submodule of the note module, reminder module, and statistics / assessment module, and the central processing module communicates with each submodule or between each submodule; the central processing module aggregates and analyzes data information; the note module is used for corporate legal management personnel to effectively record the content of legal work within a preset time unit; the reminder module is used to remind corporate legal management personnel of the content of legal work that needs to be completed according to the preset time unit; the statistical assessment module conducts corresponding assessments on the content of legal work according to multiple factor indicators and determines the performance of legal management. However, this patent only conducts corresponding assessments on the content of legal work and determines the legal management performance through multiple factor indicators, but does not consider how to provide the uploader with legal analysis texts that are helpful for analyzing the uploaded cases, nor does it consider how to improve the efficiency of legal services.

[0004] Therefore, the present invention proposes a legal processing system based on big data and artificial intelligence. Summary of the invention

[0005] The present invention provides a legal affairs processing system based on big data and artificial intelligence, which is used to accurately obtain all similar cases of each uploaded case according to the case type of each uploaded case and a preset database, and then accurately obtain all extracted phrases of each uploaded case according to the case description text of each uploaded case and the case description text of all similar cases of the corresponding uploaded case, so as to facilitate the subsequent acquisition of the word similarity of each extracted phrase of the uploaded case, accurately obtain the word similarity of each extracted phrase of each uploaded case according to each extracted phrase of each uploaded case, and realize the quantification of the overall similarity between all extracted words in each extracted phrase of the uploaded case, and then accurately obtain all similar words of each uploaded case according to the word similarity of all extracted phrases of each uploaded case. According to all similar phrases of each uploaded case, a description node graph of each uploaded case and a description node graph of all similar cases of the corresponding uploaded case are obtained, which is convenient for calculating the case similarity value. According to the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case, the case similarity value between each uploaded case and each similar case of the corresponding uploaded case is accurately obtained, and the case similarity degree between each uploaded case and each similar case of the corresponding uploaded case is quantified. Finally, according to the case similarity value between each uploaded case and all similar cases of the corresponding uploaded case, the legal processing result of each uploaded case is obtained, which is convenient for quickly providing the uploader with a legal analysis text that is helpful for analyzing the uploaded case, thereby improving the efficiency of legal services.

[0006] The present invention provides a legal affairs processing system based on big data and artificial intelligence, comprising:

[0007] A registration module, used to register and file each uploaded case, obtain the registration and file creation result of each uploaded case, and obtain the case type of each uploaded case based on the registration and file creation result of each uploaded case;

[0008] An extraction module, for obtaining all similar cases of each uploaded case based on the case type of each uploaded case and a preset database, and obtaining all extracted phrases of each uploaded case based on the case description text of each uploaded case and the case description texts of all similar cases of the corresponding uploaded case;

[0009] A calculation module, for obtaining, based on each extracted phrase of each uploaded case, a word similarity of each extracted phrase of each uploaded case, and obtaining, based on the word similarities of all extracted phrases of each uploaded case, all similar phrases of each uploaded case;

[0010] A construction module is used to obtain a description node graph of each uploaded case and description node graphs of all similar cases of the corresponding uploaded case based on all similar phrases of each uploaded case, and obtain a case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case;

[0011] The analysis module is used to obtain the legal processing result of each uploaded case based on the case similarity value between each uploaded case and all similar cases of the corresponding uploaded case.

[0012] Preferably, the registration module of the legal affairs processing system based on big data and artificial intelligence includes:

[0013] The acquisition submodule is used to treat each case uploaded to the legal affairs processing system as an uploaded case, and to treat the sum of the number of cases in the preset database before the upload time of each uploaded case and 1 as the number of each uploaded case;

[0014] The registration submodule is used to obtain the case type of each uploaded case based on the number of each uploaded case and the document of each uploaded case.

[0015] Preferably, the registration submodule of the legal affairs processing system based on big data and artificial intelligence includes:

[0016] A registration unit, which is used to obtain the party information and case description text of each uploaded case based on AI technology and the documents of each uploaded case, and obtain the registration and filing results of each uploaded case based on the number, party information and case description text of each uploaded case;

[0017] The type identification unit is used to obtain the case type of each uploaded case based on the case description text in the registration and filing results of each uploaded case and a preset case type identification model.

[0018] Preferably, the legal processing system based on big data and artificial intelligence, the extraction module, includes:

[0019] A first extraction submodule is used to extract all cases of the same case type as each uploaded case from a preset database as cases of the same type as each uploaded case;

[0020] The second extraction submodule is used to obtain the case description texts of all similar cases of each uploaded case, and obtain all extracted phrases of each uploaded case based on the case description text of each uploaded case and the case description texts of all similar cases of the corresponding uploaded case.

[0021] Preferably, the second extraction submodule of the legal affairs processing system based on big data and artificial intelligence includes:

[0022] An extraction unit, used to filter the case description text of each uploaded case and the case description text of all similar cases corresponding to the uploaded case by non-key words, and obtain all extracted words of each uploaded case and all extracted words of all similar cases corresponding to the uploaded case;

[0023] The combination unit is used to obtain all replacement words for each extracted word of each uploaded case based on a preset word library and each extracted word of each uploaded case, and to take the combination of all extracted words of each uploaded case and all extracted words of all similar cases of the corresponding uploaded case as the combined phrase of the corresponding uploaded case, and to take the combination of all extracted words in the combined phrase of each uploaded case that are repeated with all replacement words for each extracted word of the corresponding uploaded case as an extracted phrase of the corresponding uploaded case, so as to obtain all extracted phrases of each uploaded case.

[0024] Preferably, the legal affairs processing system based on big data and artificial intelligence, the computing module, includes:

[0025] A first calculation submodule, for obtaining a word similarity of each extracted phrase of each uploaded case based on each extracted phrase of each uploaded case;

[0026] The determination submodule is used to treat the corresponding extracted phrase of the corresponding uploaded case as a similar phrase of the corresponding uploaded case when the word similarity of each extracted phrase of each uploaded case is greater than a preset word similarity threshold.

[0027] Preferably, the first computing submodule of the legal affairs processing system based on big data and artificial intelligence includes:

[0028] a vector representation determination unit, used to regard the quotient between the number of words of each extracted word of each extracted phrase of each uploaded case and the number of words of all extracted words of the corresponding uploaded case as the first element of the corresponding extracted word of the corresponding extracted phrase of the corresponding uploaded case, and regard the quotient between the number of words of each extracted word of each extracted phrase of each uploaded case and the number of words of all extracted words of the corresponding extracted phrase of the corresponding uploaded case as the second element of the corresponding extracted word of the corresponding extracted phrase of the corresponding uploaded case, and regard the quotient between the first element and the second element of each extracted word of each extracted phrase of each uploaded case as the numerical representation of the corresponding extracted word of the corresponding extracted phrase of the corresponding uploaded case;

[0029] The first calculation unit is used to obtain the word similarity of each extracted phrase of each uploaded case based on the numerical representation of all extracted words of each extracted phrase of each uploaded case, that is:

[0030]

[0031] Among them, α is the word similarity of the currently calculated extracted phrase of the uploaded case, C min is the minimum value of all the extracted words in the currently calculated extracted phrase group of the currently calculated uploaded case, C max is the maximum value of the numerical representation of all the extracted words in the currently calculated extracted phrase of the currently calculated uploaded case, C0 is the mean value of the numerical representation of all the extracted words in the currently calculated extracted phrase of the currently calculated uploaded case, ln is the natural logarithm, and the value of the natural constant e is 2.718.

[0032] Preferably, the legal affairs processing system based on big data and artificial intelligence has building blocks including:

[0033] The first construction submodule is used to use the extracted word with the least number of words in each similar phrase of each uploaded case as the selected word of the corresponding similar phrase of the corresponding uploaded case, and replace all the extracted words that are repeated with each similar phrase of the corresponding uploaded case in the case description text of each uploaded case and the case description text of all similar cases of the corresponding uploaded case with the selected word of the corresponding similar phrase of the corresponding uploaded case, so as to obtain the case processing text of the corresponding uploaded case and the case processing text of all similar cases of the corresponding uploaded case;

[0034] The second construction submodule is used to obtain all extracted entities and all extracted relationships of the case processing text corresponding to the uploaded case and all extracted entities and all extracted relationships of the case processing text of all similar cases corresponding to the uploaded case based on the case processing text of each uploaded case, the case processing text of all similar cases corresponding to the uploaded case and the preset large language model, take all extracted entities of the case processing text of each uploaded case as nodes, take all extracted relationships of the case processing text of the corresponding uploaded case as relationship edges, obtain a description node graph of each uploaded case, and take all extracted entities of the case processing text of each similar case of each uploaded case as nodes, take all extracted relationships of the case processing text of the corresponding similar cases of the corresponding uploaded case as relationship edges, and obtain a description node graph of each similar case of each uploaded case;

[0035] The second calculation submodule is used to obtain the case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case.

[0036] Preferably, the second computing submodule of the legal affairs processing system based on big data and artificial intelligence includes:

[0037] A first preprocessing unit is used to obtain the node properties of all nodes and the types of all relationship edges of the description node graph of each uploaded case, and obtain the node properties of all nodes and the types of all relationship edges of the description node graph of all similar cases of each uploaded case, and use any node of the description node graph of each uploaded case and any node of the description node graph of each similar case of the corresponding uploaded case as a node group between each uploaded case and a corresponding similar case of the corresponding uploaded case, so as to obtain all node groups between each uploaded case and the corresponding similar case of the corresponding uploaded case;

[0038] A second preprocessing unit is used for treating the corresponding node group as a similar node group between the corresponding uploaded case and each similar case of the corresponding uploaded case, and treating the remaining node groups except the similar node groups between each uploaded case and each similar case of the corresponding uploaded case as non-similar node groups between the corresponding uploaded case and the corresponding similar case of the corresponding uploaded case, when the node properties of the two nodes in each node group between each uploaded case and each similar case of the corresponding uploaded case are the same and the two nodes have the same type of relationship edge.

[0039] The second calculation unit is used to obtain the case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on all similar node groups and all non-similar node groups between each uploaded case and each similar case of the corresponding uploaded case, that is:

[0040]

[0041] Where, δ is the case similarity value between the currently calculated uploaded case and the currently calculated similar case of the corresponding uploaded case, v1 is the total number of all similar node groups between the currently calculated uploaded case and the currently calculated similar case of the corresponding uploaded case, V1 is the total number of all nodes in the description node graph of the currently calculated uploaded case, V2 is the total number of all non-similar node groups between the currently calculated uploaded case and the currently calculated similar case of the corresponding uploaded case, b max is the maximum number of relationship edges among all nodes in the description node graph of the uploaded case currently calculated, b min is the minimum value of the number of relationship edges among all nodes in the description node graph of the uploaded case currently calculated, B is the mean value of the number of relationship edges among all nodes in the description node graph of the uploaded case currently calculated, k max k is the maximum number of relationship edges among all nodes in the description node graph of the currently calculated uploaded case and the currently calculated similar case, minis the minimum number of relationship edges among all nodes in the description node graph of the currently calculated uploaded case and the currently calculated similar cases, K is the mean number of relationship edges among all nodes in the description node graph of the currently calculated uploaded case and the currently calculated similar cases, ln is the natural logarithm, and the value of the natural constant e is 2.718.

[0042] Preferably, the legal affairs processing system based on big data and artificial intelligence, the analysis module includes:

[0043] A first analysis submodule is used to treat the corresponding similar case of the corresponding uploaded case as a reference case of the corresponding uploaded case when the case similarity value between each uploaded case and each similar case of the corresponding uploaded case is greater than a preset case similarity threshold;

[0044] The second analysis submodule is used to push the legal analysis texts of all reference cases of each uploaded case to the uploader's backend of the corresponding uploaded case, and obtain a new registration and filing result of the corresponding uploaded case based on the legal analysis texts of all reference cases of each uploaded case.

[0045] The beneficial effects of the present invention compared with the prior art are as follows: according to the case type of each uploaded case and a preset database, all similar cases of each uploaded case are accurately obtained, and then according to the case description text of each uploaded case and the case description texts of all similar cases of the corresponding uploaded case, all extracted phrases of each uploaded case are accurately obtained, which is convenient for subsequent acquisition of the word similarity of each extracted phrase of the uploaded case, and according to each extracted phrase of each uploaded case, the word similarity of each extracted phrase of each uploaded case is accurately obtained, which realizes quantification of the overall similarity between all extracted words in each extracted phrase of the uploaded case, and then according to the word similarity of all extracted phrases of each uploaded case, all similar phrases of each uploaded case are accurately obtained, and according to All similar phrases of each uploaded case are obtained, and the description node graph of each uploaded case and the description node graph of all similar cases of the corresponding uploaded case are obtained, which is convenient for calculating the case similarity value. According to the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case, the case similarity value between each uploaded case and each similar case of the corresponding uploaded case is accurately obtained, and the case similarity degree between each uploaded case and each similar case of the corresponding uploaded case is quantified. Finally, according to the case similarity value between each uploaded case and all similar cases of the corresponding uploaded case, the legal processing result of each uploaded case is obtained, which is convenient for quickly providing the uploader with a legal analysis text that is helpful for analyzing the uploaded case, thereby improving the efficiency of legal services.

[0046] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structure specifically pointed out in the written application document.

[0047] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0049] Figure 1 Schematic diagram of a legal affairs processing system based on big data and artificial intelligence in an embodiment of the present invention;

[0050] Figure 2 It is a specific schematic diagram of the registration module in the embodiment of the present invention. DETAILED DESCRIPTION

[0051] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0052] Embodiment 1:

[0053] The present invention provides a legal affairs processing system based on big data and artificial intelligence. Figure 1 ,include:

[0054] A registration module, used to register and file each uploaded case, obtain the registration and file creation result of each uploaded case, and obtain the case type of each uploaded case based on the registration and file creation result of each uploaded case;

[0055] An extraction module, for obtaining all similar cases of each uploaded case based on the case type of each uploaded case and a preset database, and obtaining all extracted phrases of each uploaded case based on the case description text of each uploaded case and the case description texts of all similar cases of the corresponding uploaded case;

[0056] A calculation module, for obtaining, based on each extracted phrase of each uploaded case, a word similarity of each extracted phrase of each uploaded case, and obtaining, based on the word similarities of all extracted phrases of each uploaded case, all similar phrases of each uploaded case;

[0057] A construction module is used to obtain a description node graph of each uploaded case and description node graphs of all similar cases of the corresponding uploaded case based on all similar phrases of each uploaded case, and obtain a case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case;

[0058] The analysis module is used to obtain the legal processing result of each uploaded case based on the case similarity value between each uploaded case and all similar cases of the corresponding uploaded case.

[0059] In this embodiment, the legal processing system is a legal management platform based on computer science and information technology. It obtains the legal processing results of uploaded cases in a digital and automated manner, facilitates the rapid acquisition of all legal analysis texts that are helpful for analyzing uploaded cases, and improves the efficiency of legal services.

[0060] In this embodiment, the uploaded case is a case uploaded to the legal processing system, and the case uploaded by the uploader to the legal processing system is in the form of a document.

[0061] In this embodiment, registration and filing is a process of registering each uploaded case and creating a file.

[0062] In this embodiment, the registration and filing result is to use the number of each uploaded case as the filing number, and the party information and case description text of the corresponding uploaded case as the file content.

[0063] In this embodiment, the case type is the type of each uploaded case, such as a criminal case, a labor dispute, etc.

[0064] In this embodiment, the preset database is a database composed of a large number of cases of different types collected in advance, and the collected cases are all cases with legal analysis texts.

[0065] In this embodiment, all the similar cases of each uploaded case are all the cases in the preset database that are of the same case type as each uploaded case.

[0066] In this embodiment, the case description text is a text that can reflect the case situation of the uploaded case or each case of the same type as the uploaded case.

[0067] In this embodiment, the word similarity of each extracted phrase of the uploaded case is a numerical value obtained based on each extracted phrase of the uploaded case and can reflect the overall similarity between all extracted words in each extracted phrase of the uploaded case.

[0068] In this embodiment, all similar phrases of the uploaded case are partial extracted phrases in all extracted phrases of the uploaded case, in which all extracted words in the extracted phrases are replaceable with each other.

[0069] In this embodiment, the node graph is described as a graph structure model for representing and analyzing key concepts (nodes) and their mutual relationships (relationship edges) in an uploaded case or each case of the same type as the uploaded case.

[0070] In this embodiment, the case similarity value is a numerical value obtained based on the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case, which can reflect the case similarity between each uploaded case and each similar case of the corresponding uploaded case.

[0071] In this embodiment, the legal processing result of the uploaded case is to push the legal analysis texts of all reference cases of the uploaded case to the uploader's backend of the corresponding uploaded case, and obtain a new registration and filing result of the corresponding uploaded case.

[0072] The beneficial effects of the above technology are as follows: according to the case type of each uploaded case and the preset database, all similar cases of each uploaded case are accurately obtained, and then according to the case description text of each uploaded case and the case description text of all similar cases of the corresponding uploaded case, all extracted phrases of each uploaded case are accurately obtained, which is convenient for subsequent acquisition of the word similarity of each extracted phrase of the uploaded case, and according to each extracted phrase of each uploaded case, the word similarity of each extracted phrase of each uploaded case is accurately obtained, which realizes quantification of the overall similarity between all extracted words in each extracted phrase of the uploaded case, and then according to the word similarity of all extracted phrases of each uploaded case, all similar phrases of each uploaded case are accurately obtained, and according to each uploaded case, All similar phrases of the case are obtained, and the description node graph of each uploaded case and the description node graph of all similar cases of the corresponding uploaded case are obtained, which is convenient for calculating the case similarity value. According to the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case, the case similarity value between each uploaded case and each similar case of the corresponding uploaded case is accurately obtained, and the case similarity between each uploaded case and each similar case of the corresponding uploaded case is quantified. Finally, according to the case similarity value between each uploaded case and all similar cases of the corresponding uploaded case, the legal processing result of each uploaded case is obtained, which is convenient for quickly providing the uploader with a legal analysis text that is helpful for analyzing the uploaded case, thereby improving the efficiency of legal services.

[0073] Embodiment 2:

[0074] Based on Example 1, the legal affairs processing system based on big data and artificial intelligence, the registration module, reference Figure 2 ,include:

[0075] The acquisition submodule is used to treat each case uploaded to the legal affairs processing system as an uploaded case, and to treat the sum of the number of cases in the preset database before the upload time of each uploaded case and 1 as the number of each uploaded case;

[0076] The registration submodule is used to obtain the case type of each uploaded case based on the number of each uploaded case and the document of each uploaded case.

[0077] In this embodiment, the document of the uploaded case is the document of the case uploaded by the uploader to the legal affairs processing system.

[0078] The beneficial effects of the above technology are: a specific method for determining the number of an uploaded case is given in detail, and the case type of each uploaded case is obtained according to the number of each uploaded case and the document of each uploaded case.

[0079] Embodiment 3:

[0080] On the basis of Example 2, the legal affairs processing system based on big data and artificial intelligence, the registration submodule, includes:

[0081] A registration unit, which is used to obtain the party information and case description text of each uploaded case based on AI technology and the documents of each uploaded case, and obtain the registration and filing results of each uploaded case based on the number, party information and case description text of each uploaded case;

[0082] The type identification unit is used to obtain the case type of each uploaded case based on the case description text in the registration and filing results of each uploaded case and a preset case type identification model.

[0083] In this embodiment, the AI ​​technology is an existing AI technology for automatically identifying documents of uploaded cases and obtaining the party information and case description text of the uploaded cases, such as OCR (optical character recognition) technology.

[0084] In this embodiment, the preset case type recognition model uses a large amount of case description texts collected in advance as model input, and the case type of the case corresponding to the case description text that has been manually annotated as the model output. The trained model can input case description text and output the case type of the case corresponding to the case description text.

[0085] The beneficial effects of the above technology are: according to the number of each uploaded case and the document of each uploaded case, the registration and filing results of each uploaded case are obtained, and then according to the registration and filing results of each uploaded case and the preset case type identification model, the case type of each uploaded case is obtained, which is convenient for subsequent acquisition of similar cases of the uploaded case.

[0086] Embodiment 4:

[0087] On the basis of Example 1, the legal affairs processing system based on big data and artificial intelligence, the extraction module, includes:

[0088] A first extraction submodule is used to extract all cases of the same case type as each uploaded case from a preset database as cases of the same type as each uploaded case;

[0089] The second extraction submodule is used to obtain the case description texts of all similar cases of each uploaded case, and obtain all extracted phrases of each uploaded case based on the case description text of each uploaded case and the case description texts of all similar cases of the corresponding uploaded case.

[0090] The beneficial effects of the above technology are: according to the case type of each uploaded case and the preset database, all similar cases of each uploaded case are accurately obtained, and all similar cases of each uploaded case are quickly and accurately matched, avoiding the tedious and time-consuming manual matching, and improving the efficiency of case processing. Then, according to the case description text of each uploaded case and the case description text of all similar cases of the corresponding uploaded case, all extracted phrases of each uploaded case are accurately obtained, which facilitates the subsequent acquisition of the word similarity of each extracted phrase of the uploaded case.

[0091] Embodiment 5:

[0092] On the basis of Example 4, the legal affairs processing system based on big data and artificial intelligence, the second extraction submodule includes:

[0093] An extraction unit, used to filter the case description text of each uploaded case and the case description text of all similar cases corresponding to the uploaded case by non-key words, and obtain all extracted words of each uploaded case and all extracted words of all similar cases corresponding to the uploaded case;

[0094] The combination unit is used to obtain all replacement words for each extracted word of each uploaded case based on a preset word library and each extracted word of each uploaded case, and to take the combination of all extracted words of each uploaded case and all extracted words of all similar cases of the corresponding uploaded case as the combined phrase of the corresponding uploaded case, and to take the combination of all extracted words in the combined phrase of each uploaded case that are repeated with all replacement words for each extracted word of the corresponding uploaded case as an extracted phrase of the corresponding uploaded case, so as to obtain all extracted phrases of each uploaded case.

[0095] In this embodiment, non-key words are words in the case description text that do not contribute substantially to the semantic understanding or sentiment analysis of the case situation, such as stop words.

[0096] In this embodiment, filtering is a process of deleting non-key words from the case description text of each uploaded case and the case description texts of all similar cases corresponding to the uploaded case.

[0097] In this embodiment, all extracted words of the uploaded case are all words after filtering the case description text of each uploaded case for non-key words.

[0098] In this embodiment, the preset word library is a word library composed of all extracted words of a large number of cases collected in advance.

[0099] In this embodiment, the replacement word is a word selected from a preset word library that can replace each extracted word of each uploaded case (synonyms and near synonyms of each extracted word of each uploaded case), and the replacement word of the extracted word includes the corresponding extracted word itself.

[0100] In this embodiment, the combined phrase of the uploaded case is a combination of all the extracted words of the uploaded case and all the extracted words of all similar cases corresponding to the uploaded case, and there is no duplication of the extracted words in the combined phrase.

[0101] The beneficial effects of the above technology are: according to the case description text of each uploaded case and the case description text of all similar cases corresponding to the uploaded case, all extracted words of each uploaded case and all extracted words of each similar case of the corresponding uploaded case are obtained, and then according to all extracted words of each uploaded case and all extracted words of each similar case of the corresponding uploaded case, all extracted phrases of each uploaded case are accurately obtained, which is convenient for subsequent acquisition of the word similarity of each extracted phrase of the uploaded case. This embodiment gives in detail a specific method for accurately obtaining all extracted phrases of each uploaded case according to the case description text of each uploaded case and the case description text of all similar cases corresponding to the uploaded case.

[0102] Embodiment 6:

[0103] On the basis of Example 1, the legal affairs processing system based on big data and artificial intelligence, the computing module, includes:

[0104] A first calculation submodule, for obtaining a word similarity of each extracted phrase of each uploaded case based on each extracted phrase of each uploaded case;

[0105] The determination submodule is used to treat the corresponding extracted phrase of the corresponding uploaded case as a similar phrase of the corresponding uploaded case when the word similarity of each extracted phrase of each uploaded case is greater than a preset word similarity threshold.

[0106] In this embodiment, the preset word similarity threshold is a preset word similarity threshold for obtaining similar word groups of the uploaded case.

[0107] The beneficial effect of the above technology is: according to the word similarity of all extracted phrases of each uploaded case and the preset word similarity threshold, all similar phrases of each uploaded case are accurately obtained, which is convenient for the construction of the subsequent description node graph.

[0108] Embodiment 7:

[0109] On the basis of Example 6, the legal affairs processing system based on big data and artificial intelligence, the first computing submodule includes:

[0110] a vector representation determination unit, used to regard the quotient between the number of words of each extracted word of each extracted phrase of each uploaded case and the number of words of all extracted words of the corresponding uploaded case as the first element of the corresponding extracted word of the corresponding extracted phrase of the corresponding uploaded case, and regard the quotient between the number of words of each extracted word of each extracted phrase of each uploaded case and the number of words of all extracted words of the corresponding extracted phrase of the corresponding uploaded case as the second element of the corresponding extracted word of the corresponding extracted phrase of the corresponding uploaded case, and regard the quotient between the first element and the second element of each extracted word of each extracted phrase of each uploaded case as the numerical representation of the corresponding extracted word of the corresponding extracted phrase of the corresponding uploaded case;

[0111] The first calculation unit is used to obtain the word similarity of each extracted phrase of each uploaded case based on the numerical representation of all extracted words of each extracted phrase of each uploaded case, that is:

[0112]

[0113] Among them, α is the word similarity of the currently calculated extracted phrase of the uploaded case, C min is the minimum value of all the extracted words in the currently calculated extracted phrase group of the currently calculated uploaded case, C max is the maximum value of the numerical representation of all the extracted words in the currently calculated extracted phrase of the currently calculated uploaded case, C0 is the mean value of the numerical representation of all the extracted words in the currently calculated extracted phrase of the currently calculated uploaded case, ln is the natural logarithm, and the value of the natural constant e is 2.718.

[0114] The beneficial effect of the above technology is: according to each extracted phrase of each uploaded case, the numerical representation of all extracted words of each extracted phrase of each uploaded case is obtained, and then according to the numerical representation of all extracted words of each extracted phrase of each uploaded case, the word similarity of each extracted phrase of each uploaded case is obtained. This embodiment gives in detail a specific method for obtaining the word similarity of each extracted phrase of each uploaded case according to each extracted phrase of each uploaded case.

[0115] Embodiment 8:

[0116] On the basis of Example 1, a legal affairs processing system based on big data and artificial intelligence is constructed with modules including:

[0117] The first construction submodule is used to use the extracted word with the least number of words in each similar phrase of each uploaded case as the selected word of the corresponding similar phrase of the corresponding uploaded case, and replace all the extracted words that are repeated with each similar phrase of the corresponding uploaded case in the case description text of each uploaded case and the case description text of all similar cases of the corresponding uploaded case with the selected word of the corresponding similar phrase of the corresponding uploaded case, so as to obtain the case processing text of the corresponding uploaded case and the case processing text of all similar cases of the corresponding uploaded case;

[0118] The second construction submodule is used to obtain all extracted entities and all extracted relationships of the case processing text corresponding to the uploaded case and all extracted entities and all extracted relationships of the case processing text of all similar cases corresponding to the uploaded case based on the case processing text of each uploaded case, the case processing text of all similar cases corresponding to the uploaded case and the preset large language model, take all extracted entities of the case processing text of each uploaded case as nodes, take all extracted relationships of the case processing text of the corresponding uploaded case as relationship edges, obtain a description node graph of each uploaded case, and take all extracted entities of the case processing text of each similar case of each uploaded case as nodes, take all extracted relationships of the case processing text of the corresponding similar cases of the corresponding uploaded case as relationship edges, and obtain a description node graph of each similar case of each uploaded case;

[0119] The second calculation submodule is used to obtain the case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case.

[0120] In this embodiment, the case processing text is a text obtained by replacing part of the extracted words in the case description text of the uploaded case or the case description text of the same type of case as the uploaded case.

[0121] In this embodiment, the preset large language model is an existing model that is pre-set to extract all extracted entities (such as names of people, places, organization names, etc.) and all extracted relations (such as "born in", "belongs to", etc.) in the case processing text, such as GPT, BERT, etc.

[0122] The beneficial effects of the above technology are: according to all similar phrases of each uploaded case, the case processing text of each uploaded case and the case processing text of all similar cases of the corresponding uploaded case are obtained, and then according to the case processing text of the uploaded case and the case processing text of all similar cases of the corresponding uploaded case, the description node graph of each uploaded case and the description node graph of all similar cases of the corresponding uploaded case are obtained, which is convenient for calculating the case similarity value, and according to the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case, the case similarity value between each uploaded case and each similar case of the corresponding uploaded case is accurately obtained. This embodiment gives in detail a specific method for obtaining the description node graph of each uploaded case and the description node graph of all similar cases of the corresponding uploaded case according to the case processing text of the uploaded case and the case processing text of all similar cases of the corresponding uploaded case.

[0123] Embodiment 9:

[0124] On the basis of Example 8, the legal affairs processing system based on big data and artificial intelligence, the second computing submodule includes:

[0125] A first preprocessing unit is used to obtain the node properties of all nodes and the types of all relationship edges of the description node graph of each uploaded case, and obtain the node properties of all nodes and the types of all relationship edges of the description node graph of all similar cases of each uploaded case, and use any node of the description node graph of each uploaded case and any node of the description node graph of each similar case of the corresponding uploaded case as a node group between each uploaded case and a corresponding similar case of the corresponding uploaded case, so as to obtain all node groups between each uploaded case and the corresponding similar case of the corresponding uploaded case;

[0126] A second preprocessing unit is used for treating the corresponding node group as a similar node group between the corresponding uploaded case and each similar case of the corresponding uploaded case, and treating the remaining node groups except the similar node groups between each uploaded case and each similar case of the corresponding uploaded case as non-similar node groups between the corresponding uploaded case and the corresponding similar case of the corresponding uploaded case, when the node properties of the two nodes in each node group between each uploaded case and each similar case of the corresponding uploaded case are the same and the two nodes have the same type of relationship edge.

[0127] The second calculation unit is used to obtain the case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on all similar node groups and all non-similar node groups between each uploaded case and each similar case of the corresponding uploaded case, that is:

[0128]

[0129] Where, δ is the case similarity value between the currently calculated uploaded case and the currently calculated similar case of the corresponding uploaded case, v1 is the total number of all similar node groups between the currently calculated uploaded case and the currently calculated similar case of the corresponding uploaded case, V1 is the total number of all nodes in the description node graph of the currently calculated uploaded case, V2 is the total number of all non-similar node groups between the currently calculated uploaded case and the currently calculated similar case of the corresponding uploaded case, b max is the maximum number of relationship edges among all nodes in the description node graph of the uploaded case currently calculated, b min is the minimum value of the number of relationship edges among all nodes in the description node graph of the uploaded case currently calculated, B is the mean value of the number of relationship edges among all nodes in the description node graph of the uploaded case currently calculated, k max k is the maximum number of relationship edges among all nodes in the description node graph of the currently calculated uploaded case and the currently calculated similar case, min is the minimum number of relationship edges among all nodes in the description node graph of the currently calculated uploaded case and the currently calculated similar cases, K is the mean number of relationship edges among all nodes in the description node graph of the currently calculated uploaded case and the currently calculated similar cases, ln is the natural logarithm, and the value of the natural constant e is 2.718.

[0130] In this embodiment, the node property is a property that describes a node of the node graph, such as a person, a place, etc.

[0131] In this embodiment, the type of the relationship edge is the type of the relationship edge on the node describing the node graph, such as a time relationship, a causal relationship, and the like.

[0132] The beneficial effects of the above technology are: based on the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case, all similar node groups and all non-similar node groups between each uploaded case and each similar case of the corresponding uploaded case are obtained, and then based on all similar node groups and all non-similar node groups between each uploaded case and each similar case of the corresponding uploaded case, the case similarity value between each uploaded case and each similar case of the corresponding uploaded case is accurately obtained, thereby realizing the quantification of the case similarity between each uploaded case and each similar case of the corresponding uploaded case.

[0133] Embodiment 10:

[0134] On the basis of Example 1, the legal affairs processing system based on big data and artificial intelligence, the analysis module, includes:

[0135] A first analysis submodule is used to treat the corresponding similar case of the corresponding uploaded case as a reference case of the corresponding uploaded case when the case similarity value between each uploaded case and each similar case of the corresponding uploaded case is greater than a preset case similarity threshold;

[0136] The second analysis submodule is used to push the legal analysis texts of all reference cases of each uploaded case to the uploader's backend of the corresponding uploaded case, and obtain a new registration and filing result of the corresponding uploaded case based on the legal analysis texts of all reference cases of each uploaded case.

[0137] In this embodiment, the preset case similarity threshold is a case similarity threshold that is preset to obtain a reference case for the uploaded case.

[0138] In this embodiment, the reference case of the uploaded case is a case that is helpful for analyzing the uploaded case.

[0139] In this embodiment, the legal analysis text is a text formed after the lawyer conducts a legal analysis of the reference case.

[0140] In this embodiment, the uploader's backend is the backend for the uploader to log into the legal processing system.

[0141] In this embodiment, the new registration and filing result of the uploaded case is to use the legal analysis texts of all reference cases of each uploaded case as the new file content of the registration and filing result of the uploaded case.

[0142] The beneficial effects of the above technology are: based on the case similarity value between each uploaded case and all similar cases corresponding to the uploaded case and the preset case similarity threshold, the legal processing result of each uploaded case is obtained, which is convenient for quickly providing the uploader with a legal analysis text that is helpful for analyzing the uploaded case, thereby improving the efficiency of legal services.

[0143] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention, and the present invention is also intended to include these changes and modifications.

Claims

1. A legal affairs processing system based on big data and artificial intelligence, characterized in that: include: A registration module, used to register and file each uploaded case, obtain the registration and file creation result of each uploaded case, and obtain the case type of each uploaded case based on the registration and file creation result of each uploaded case; An extraction module, for obtaining all similar cases of each uploaded case based on the case type of each uploaded case and a preset database, and obtaining all extracted phrases of each uploaded case based on the case description text of each uploaded case and the case description texts of all similar cases of the corresponding uploaded case; A calculation module, for obtaining, based on each extracted phrase of each uploaded case, a word similarity of each extracted phrase of each uploaded case, and obtaining, based on the word similarities of all extracted phrases of each uploaded case, all similar phrases of each uploaded case; A construction module is used to obtain a description node graph of each uploaded case and description node graphs of all similar cases of the corresponding uploaded case based on all similar phrases of each uploaded case, and obtain a case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case; An analysis module, for obtaining a legal processing result for each uploaded case based on a case similarity value between each uploaded case and all similar cases of the corresponding uploaded case; The computing module includes: A first calculation submodule, for obtaining a word similarity of each extracted phrase of each uploaded case based on each extracted phrase of each uploaded case; A determination submodule, for treating the corresponding extracted phrase of the corresponding uploaded case as a similar phrase of the corresponding uploaded case when the word similarity of each extracted phrase of each uploaded case is greater than a preset word similarity threshold; The first computing submodule includes: a vector representation determination unit, used to regard the quotient between the number of words of each extracted word of each extracted phrase of each uploaded case and the number of words of all extracted words of the corresponding uploaded case as the first element of the corresponding extracted word of the corresponding extracted phrase of the corresponding uploaded case, and regard the quotient between the number of words of each extracted word of each extracted phrase of each uploaded case and the number of words of all extracted words of the corresponding extracted phrase of the corresponding uploaded case as the second element of the corresponding extracted word of the corresponding extracted phrase of the corresponding uploaded case, and regard the quotient between the first element and the second element of each extracted word of each extracted phrase of each uploaded case as the numerical representation of the corresponding extracted word of the corresponding extracted phrase of the corresponding uploaded case; The first calculation unit is used to obtain the word similarity of each extracted phrase of each uploaded case based on the numerical representation of all extracted words of each extracted phrase of each uploaded case, that is: Among them, α is the word similarity of the currently calculated extracted phrase of the uploaded case, C min is the minimum value of all the extracted words in the extracted phrase group of the uploaded case currently being calculated, C max is the maximum value of the numerical representation of all the extraction words of the currently calculated extraction phrase of the currently calculated uploaded case, C0 is the mean value of the numerical representation of all the extraction words of the currently calculated extraction phrase of the currently calculated uploaded case, ln is the natural logarithm, and the value of the natural constant e is 2.718; Among them, the building blocks include: The first construction submodule is used to use the extracted word with the least number of words in each similar phrase of each uploaded case as the selected word of the corresponding similar phrase of the corresponding uploaded case, and replace all the extracted words that are repeated with each similar phrase of the corresponding uploaded case in the case description text of each uploaded case and the case description text of all similar cases of the corresponding uploaded case with the selected word of the corresponding similar phrase of the corresponding uploaded case, so as to obtain the case processing text of the corresponding uploaded case and the case processing text of all similar cases of the corresponding uploaded case; The second construction submodule is used to obtain all extracted entities and all extracted relationships of the case processing text corresponding to the uploaded case and all extracted entities and all extracted relationships of the case processing text of all similar cases corresponding to the uploaded case based on the case processing text of each uploaded case, the case processing text of all similar cases corresponding to the uploaded case and the preset large language model, take all extracted entities of the case processing text of each uploaded case as nodes, take all extracted relationships of the case processing text of the corresponding uploaded case as relationship edges, obtain a description node graph of each uploaded case, and take all extracted entities of the case processing text of each similar case of each uploaded case as nodes, take all extracted relationships of the case processing text of the corresponding similar cases of the corresponding uploaded case as relationship edges, and obtain a description node graph of each similar case of each uploaded case; A second calculation submodule is used to obtain a case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on the description node graph of each uploaded case and the description node graph of each similar case of the corresponding uploaded case; The second computing submodule includes: A first preprocessing unit is used to obtain the node properties of all nodes and the types of all relationship edges of the description node graph of each uploaded case, and obtain the node properties of all nodes and the types of all relationship edges of the description node graph of all similar cases of each uploaded case, and use any node of the description node graph of each uploaded case and any node of the description node graph of each similar case of the corresponding uploaded case as a node group between each uploaded case and a corresponding similar case of the corresponding uploaded case, so as to obtain all node groups between each uploaded case and the corresponding similar case of the corresponding uploaded case; A second preprocessing unit is used for treating the corresponding node group as a similar node group between the corresponding uploaded case and each similar case of the corresponding uploaded case, and treating the remaining node groups except the similar node groups between each uploaded case and each similar case of the corresponding uploaded case as non-similar node groups between the corresponding uploaded case and the corresponding similar case of the corresponding uploaded case, when the node properties of the two nodes in each node group between each uploaded case and each similar case of the corresponding uploaded case are the same and the two nodes have the same type of relationship edge. The second calculation unit is used to obtain the case similarity value between each uploaded case and each similar case of the corresponding uploaded case based on all similar node groups and all non-similar node groups between each uploaded case and each similar case of the corresponding uploaded case, that is: Where, δ is the case similarity value between the currently calculated uploaded case and the currently calculated similar case of the corresponding uploaded case, v1 is the total number of all similar node groups between the currently calculated uploaded case and the currently calculated similar case of the corresponding uploaded case, V1 is the total number of all nodes in the description node graph of the currently calculated uploaded case, V2 is the total number of all non-similar node groups between the currently calculated uploaded case and the currently calculated similar case of the corresponding uploaded case, b max is the maximum number of relationship edges among all nodes in the description node graph of the uploaded case currently calculated, b min is the minimum value of the number of relationship edges among all nodes in the description node graph of the uploaded case currently calculated, B is the mean value of the number of relationship edges among all nodes in the description node graph of the uploaded case currently calculated, k max k is the maximum number of relationship edges among all nodes in the description node graph of the currently calculated uploaded case and the currently calculated similar case, min is the minimum number of relationship edges among all nodes in the description node graph of the currently calculated uploaded case and the currently calculated similar cases, K is the mean number of relationship edges among all nodes in the description node graph of the currently calculated uploaded case and the currently calculated similar cases, ln is the natural logarithm, and the value of the natural constant e is 2.

718.

2. The legal affairs processing system based on big data and artificial intelligence according to claim 1 is characterized in that: Registration module, including: The acquisition submodule is used to treat each case uploaded to the legal affairs processing system as an uploaded case, and treat the sum of the number of cases in the preset database before the upload time of each uploaded case and 1 as the number of each uploaded case; The registration submodule is used to obtain the case type of each uploaded case based on the number of each uploaded case and the document of each uploaded case.

3. The legal affairs processing system based on big data and artificial intelligence according to claim 2 is characterized in that: Registration submodule, including: A registration unit, which is used to obtain the party information and case description text of each uploaded case based on AI technology and the documents of each uploaded case, and obtain the registration and filing results of each uploaded case based on the number, party information and case description text of each uploaded case; The type identification unit is used to obtain the case type of each uploaded case based on the case description text in the registration and filing results of each uploaded case and a preset case type identification model.

4. The legal affairs processing system based on big data and artificial intelligence according to claim 1 is characterized in that: Extraction modules, including: A first extraction submodule is used to extract all cases of the same case type as each uploaded case from a preset database as cases of the same type as each uploaded case; The second extraction submodule is used to obtain the case description texts of all similar cases of each uploaded case, and obtain all extracted phrases of each uploaded case based on the case description text of each uploaded case and the case description texts of all similar cases of the corresponding uploaded case.

5. The legal affairs processing system based on big data and artificial intelligence according to claim 4 is characterized in that: The second extraction submodule includes: An extraction unit, used to filter the case description text of each uploaded case and the case description text of all similar cases corresponding to the uploaded case by non-key words, and obtain all extracted words of each uploaded case and all extracted words of all similar cases corresponding to the uploaded case; The combination unit is used to obtain all replacement words for each extracted word of each uploaded case based on a preset word library and each extracted word of each uploaded case, and to take the combination of all extracted words of each uploaded case and all extracted words of all similar cases of the corresponding uploaded case as the combined phrase of the corresponding uploaded case, and to take the combination of all extracted words in the combined phrase of each uploaded case that are repeated with all replacement words for each extracted word of the corresponding uploaded case as an extracted phrase of the corresponding uploaded case, so as to obtain all extracted phrases of each uploaded case.

6. The legal affairs processing system based on big data and artificial intelligence according to claim 1 is characterized in that: Analysis modules, including: A first analysis submodule is used to treat the corresponding similar case of the corresponding uploaded case as a reference case of the corresponding uploaded case when the case similarity value between each uploaded case and each similar case of the corresponding uploaded case is greater than a preset case similarity threshold; The second analysis submodule is used to push the legal analysis texts of all reference cases of each uploaded case to the uploader's backend of the corresponding uploaded case, and obtain a new registration and filing result of the corresponding uploaded case based on the legal analysis texts of all reference cases of each uploaded case.

Images