WO2021051931A1

WO2021051931A1 - Method and device for determining probability of winning legal case, apparatus, and computer readable storage medium

Info

Publication number: WO2021051931A1
Application number: PCT/CN2020/098844
Authority: WO
Inventors: 聂宇昕; 邓俊豪; 徐冰; 陈晨; 汪伟
Original assignee: 平安科技（深圳）有限公司
Priority date: 2019-09-18
Filing date: 2020-06-29
Publication date: 2021-03-25
Also published as: CN110825880A

Abstract

A method and device for determining a probability of winning a legal case, an apparatus, and a computer readable storage medium. The method comprises: acquiring a first knowledge graph of a target case and second knowledge graphs of multiple decided cases, wherein the target case is a case for which a probability of winning is to be predicted (S101); performing calculation according to the first knowledge graph and the respective second knowledge graphs to obtain levels of similarity between the target case and the respective decided cases (S102); determining, according to the levels of similarity between the target case and the respective decided cases, decided cases similar to the target case (S103); and determining a probability of winning the target case according to the decided cases similar to the target case (S104). The method relates to data analysis and knowledge graphs, and can be used to quickly and accurately predict a probability of winning a legal case.

Description

Method, device, equipment and computer readable storage medium for determining case winning rate

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on September 18, 2019, the application number is 201910883531.2, and the invention title is "Method, Apparatus, Equipment, and Computer-readable Storage Medium for Determining Case Win Rate", and its entire contents Incorporated in this application by reference.

Technical field

This application relates to the technical field of data analysis, and in particular to a method, device, equipment and computer-readable storage medium for determining a case winning rate.

Background technique

With the development of society and economy and the improvement of the rule of law system, the people's judicial needs are increasing. When the people request the court to try a case, they need to provide petitions and evidence. Evidence is especially important and affects the outcome of the case. However, the court has a long period of hearing cases, and it takes a long time to know the results of the court’s trial, and it is impossible to know the probability of winning the case in advance. To this end, the people can consult with experienced lawyers, and the lawyers can predict the probability of success based on the information of past convicted cases and current cases. However, the inventor realizes that it takes a long time for lawyers to predict the probability of winning a pending case based on the trial results of similar cases that have already been tried subjectively, and the accuracy of the predicted probability of winning cannot be guaranteed. Therefore, how to quickly and accurately predict the probability of winning a case is a problem that needs to be solved urgently.

technical problem

The main purpose of this application is to provide a method, device, equipment, and computer-readable storage medium for determining the winning rate of a case, so as to quickly and accurately predict the probability of winning a case.

Technical solutions

In the first aspect, the present application provides a method for determining the winning rate of a case. The method for determining the winning rate of a case includes the following steps: acquiring a first knowledge graph of a target case and a second knowledge graph of multiple convicted cases, wherein the target case Is the case for which the probability of winning is to be predicted; according to the first knowledge graph and each of the second knowledge graphs, the similarity between the target case and each of the sentenced cases is calculated; according to the target case and The degree of similarity between each of the sentenced cases determines the sentenced cases that are similar to the target case; and the probability of winning the target case is determined based on the sentenced cases that are similar to the target case.

In a second aspect, the present application also provides a case winning rate determining device, the case winning rate determining device comprising: an acquisition module for acquiring a first knowledge graph of a target case and a second knowledge graph of a plurality of judged cases, wherein, The target case is a case for which the probability of winning a case is to be predicted; a calculation module is used to calculate the relationship between the target case and each of the sentenced cases according to the first knowledge graph and each of the second knowledge graphs Similarity; a determination module, used to determine a sentenced case similar to the target case based on the similarity between the target case and each of the sentenced cases; the determination module is also used to determine a case that is similar to the target case For judged cases that are similar to the target case, determine the probability of winning the target case.

In a third aspect, the present application also provides a computer device that includes a processor, a memory, and a computer program stored on the memory and executable by the processor, wherein the computer program is When the processor executes, the following steps are realized: acquiring a first knowledge graph of a target case and a second knowledge graph of multiple judged cases, wherein the target case is a case for which the probability of winning a case is to be predicted; according to the first knowledge graph And each of the second knowledge graphs, calculate the similarity between the target case and each of the sentenced cases; according to the similarity between the target case and each of the sentenced cases, determine the Sentenced cases that are similar to the target case; determine the probability of winning the target case based on the sentenced cases that are similar to the target case.

In a fourth aspect, this application also provides a computer-readable storage medium having a computer program stored on the computer-readable storage medium, and when the computer program is executed by a processor, the following steps are implemented: obtaining the first case of the target case A knowledge graph and a second knowledge graph of multiple convicted cases, wherein the target case is a case for which the probability of winning a case is to be predicted; and the target case is calculated according to the first knowledge graph and each of the second knowledge graphs The degree of similarity with each of the sentenced cases; according to the degree of similarity between the target case and each of the sentenced cases, determine the sentenced cases that are similar to the target case; For the similarly judged cases, determine the probability of winning the target case.

Beneficial effect

This application uses the knowledge map of the case to be predicted and the knowledge map of the judged case to calculate the similarity between the case to be predicted and each sentenced case, and it is based on the case to be predicted. Based on the similarity between the judged cases, determine the judged cases similar to the case, and then determine the probability of winning the case based on the sentenced cases similar to the case, which can quickly and more accurately determine the probability of winning the case.

Description of the drawings

FIG. 1 is a schematic flowchart of a method for determining a case winning rate provided by an embodiment of the application.

Fig. 2 is a schematic diagram of the knowledge graph in an embodiment of the application.

FIG. 3 is a schematic flowchart of another method for determining a case winning rate provided by an embodiment of the application.

Fig. 4 is a schematic block diagram of an apparatus for determining a case winning rate provided by an embodiment of the application.

Fig. 5 is a schematic block diagram of another device for determining a case winning rate provided by an embodiment of the application.

FIG. 6 is a schematic block diagram of the structure of a computer device related to an embodiment of this application.

Embodiments of the present invention

The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

The flowchart shown in the drawings is only an example, and does not necessarily include all contents and operations/steps, nor does it have to be executed in the described order. For example, some operations/steps can also be decomposed, combined or partially combined, so the actual execution order may be changed according to actual conditions.

The technical solution of the present application can be applied to the field of artificial intelligence or big data technology, and can be used to predict and analyze the winning rate of cases. The data involved can be stored in a database or distributed through a blockchain, which is not limited by this application.

The embodiments of the present application provide a method, a device, a computer device, and a computer-readable storage medium for determining a case winning rate. Among them, the method for determining the winning percentage of the case can be applied to a server, and the server can be a single server or a server cluster composed of multiple servers.

Hereinafter, some embodiments of the present application will be described in detail with reference to the accompanying drawings. In the case of no conflict, the following embodiments and features in the embodiments can be combined with each other.

Please refer to FIG. 1, which is a schematic flowchart of a method for determining a case winning rate provided by an embodiment of the application.

As shown in Fig. 1, the method for determining the case winning percentage includes steps S101 to S104.

Step S101: Obtain a first knowledge graph of a target case and a second knowledge graph of a plurality of judged cases, where the target case is a case for which the probability of winning the case is to be predicted.

The server can obtain the first knowledge map of the target case and the second knowledge map of multiple judged cases in real time or at regular intervals, where the target case is a case for which the probability of winning the case is to be predicted. Specifically, when receiving the case winning rate prediction request sent by the terminal device, the server obtains the case number from the case winning rate prediction request, and obtains the first knowledge graph corresponding to the case number and the second knowledge graphs of multiple judged cases .

Among them, the server stores cases for which the probability of winning is to be predicted, that is, the knowledge map of the target case, which is recorded as the first knowledge map, and the knowledge map of the judged case is also stored, recorded as the second knowledge map, and the case’s knowledge map stores cases Knowledge information and legal knowledge information. The case knowledge information includes basic case knowledge, dispute focus knowledge, factual element knowledge, and evidence knowledge. Basic case knowledge includes but is not limited to litigants, litigation companies, litigation participants, plaintiffs, defendants, Viewpoints of appeal and defense, dispute focus knowledge includes dispute focus, and evidence knowledge includes evidence items, evidence types, and evidence attributes. Evidence attribute refers to the characteristic attribute of the evidence, such as whether the IOU has the signature of the borrower. There is at least one fact in every plaintiff request, such as "A requests B to repay the money". The fact is "A lent money to B and B does not repay it."

The following takes a sentenced case as an example to explain how to construct the knowledge graph of a sentenced case. Specifically, obtain the judgment document of the judged case, and extract the case knowledge from the judgment document to obtain the structured case knowledge information and legal knowledge information of the judged case, and then according to the case knowledge information and legal knowledge information, Construct a knowledge graph of convicted cases. Among them, the ruling document includes the relationship between the characters, the origin of the case, the trial process, the facts, and the reason and basis of the judgment. The character relationship part includes the basic information of the parties, the basic information of the entrusted agents ad litem, and the litigation status of the parties; the origin part of the case includes information about the petition and plea, the part of the trial process is the court record, and the fact part includes the plaintiff’s litigation request, Facts and reasons, the facts and reasons of the defendant’s defense, the facts found by the court and the evidence on which the case was determined, the reasons and basis of the judgment include the relationship between the evidence and the focus of the dispute, and the relationship between the basis of the judgment and the legal provisions.

In one embodiment, the method for extracting case knowledge is specifically: extracting basic case knowledge from the ruling document through a basic case knowledge extraction model; extracting dispute focus knowledge from the ruling document through a dispute focus extraction model, and extracting it through evidence items The model extracts the evidence knowledge from the ruling document; the legal knowledge extraction model extracts the legal knowledge from the ruling document.

Among them, the basic case knowledge extraction model includes a text segmentation layer and an information extraction layer. The text segmentation layer is provided with a set of regular expressions for segmentation, and the information extraction layer is provided with regular expressions for extracting the opinions of the plaintiff and the petition. A collection of formulas, a collection of regular expressions used to extract the defendant’s and argued opinions, and a collection of regular expressions used to extract the relationship between litigation figures, litigation companies, and litigation participants. The basic case knowledge extraction method is specifically: through this text segmentation The regular expression set in the layer, extract the petition paragraph, the argument paragraph and the litigation participant paragraph from the ruling document, and extract the plaintiff and the petition from the petition paragraph through each regular expression set in the information extraction layer Viewpoints, extract the defendant and defense views from the defense paragraph, and extract the relationship between the litigation person, the litigation company, and the litigation participant from the litigation participant paragraph.

Among them, the evidence item extraction model includes an evidence paragraph extraction layer, an evidence item extraction layer, an evidence classification layer, and an evidence attribute determination layer. The evidence paragraph extraction layer is provided with a regular expression set for extracting evidence paragraphs, and the potential evidence item extraction layer A potential evidence item extraction model is set, the evidence classification layer is set with an evidence classification table, and the evidence attribute determination layer is set with a similarity calculation model and a mapping relationship table between evidence attributes and evidence keywords.

In one embodiment, the method of extracting evidence knowledge is specifically as follows: the server extracts a paragraph of evidence from the judgment document, and divides the paragraph of the evidence to obtain a number of evidence sentences; and extracts a model from the potential evidence item in the evidence item extraction layer. Extract the potential evidence items in each evidence sentence; determine the evidence category of each potential evidence item through the evidence classification table in the evidence classification layer, that is, query the evidence classification table, obtain the evidence categories of the potential evidence items, and calculate the potential evidence The similarity between the item and the evidence keyword corresponding to each sub-category under the major category of evidence, and then the sub-category with the largest similarity is determined as the evidence category of the potential evidence item; the level of evidence attribute determination is used to determine each category of evidence. The evidence attribute of a type of evidence, that is, for each type of evidence in the evidence paragraph, the evidence paragraph is traversed, the first evidence statement and the last evidence statement appear, and the two evidence statements and the two evidence The paragraph text between the sentences is determined as the context information of this type of evidence; query the mapping relationship table between the evidence attributes and the evidence keywords, and obtain the target evidence sentences containing the evidence keywords from the context information of each type of evidence, and obtain The evidence attribute group corresponding to the target evidence sentence, and then after the target evidence sentence and the evidence attribute group are spliced, the similarity calculation model is input to calculate the similarity between each evidence attribute in the evidence attribute group and the target evidence sentence, and The evidence attribute with the highest similarity is used as the target evidence attribute of the target evidence sentence, so as to extract all evidence, evidence categories and evidence attributes and other evidence knowledge.

It should be noted that the above keyword set, the mapping relationship table between evidence attributes and evidence keywords, and each regular expression set are set based on actual conditions, and this application does not specifically limit this. The above evidence item extraction model is obtained through the AutoNER model based on manually labeled sample data. The AutoNER model uses the word vector trained for the ruling document as the embedding. Because the accuracy of the automatic remote labeling data is too low, this module is abandoned and used instead. For training on labeled data, the accuracy of manually labeled sample data is high. In addition, in order to prevent over-fitting, the method of data enhancement is used during training, that is, random replacement of no more than 3 words and/or in the evidence sentence Reverse the order of the words in the evidence sentence. The above-mentioned similarity model is obtained by retraining the pre-training model of the BERT model based on legal corpus, reducing the encoder module (encoding module) of the BERT model to 3 layers, and adjusting the sentence length to achieve time optimization. In addition, For training tasks, obtain interference sample data, train the model based on the interference sample data, and connect the classification layer to the position of the encoder layer. The AutoNER model is a model that can automatically label data and train named entity recognition without manual labeling. The BERT model is the first deep, bidirectional and unsupervised language representation model. Optionally, the regular expression used to obtain the appeal paragraph is: r*^(?P<appeal>.*?)\n[^\n]*? ( |Statement|Not [out to] court|recognition).

Among them, the construction method of the knowledge graph is specifically as follows: the law in the knowledge of the law and the plaintiff, defendant, petition, argument, dispute focus, fact elements and evidence in the knowledge information of the case are used as the entity nodes of the knowledge graph. And from the legal knowledge and case knowledge information to obtain the relationship and attributes between each entity node (plaintiff, defendant, claim point of view, argued point of view, dispute focus, factual elements, evidence and specific value of the legal provisions), and then Based on the entity node, the relationship between the entity node and the attribute of the entity node, the case knowledge graph of the target case is constructed. Please refer to Figure 2. Figure 2 is a schematic diagram of the knowledge graph in the embodiment of the application. As shown in Figure 2, the entity nodes of the knowledge graph are the plaintiff, the defendant, the petition point of view, the argued point of view, the focus of the dispute, the fact element, Evidence and legal provisions, and factual elements include small element 1, small element 2, and small element 3.

Step S102: Calculate the similarity between the target case and each of the sentenced cases according to the first knowledge graph and each of the second knowledge graphs.

The server calculates the similarity between the target case and each sentenced case based on the first knowledge graph and the second knowledge graph of each sentenced case. Specifically, the similarity between the first knowledge graph and each second knowledge graph is calculated; the similarity between the first knowledge graph and each second knowledge graph is taken as the difference between the target case and each judged case Similarity. The knowledge graph contains the complete information of the case. By calculating the similarity between the knowledge graphs, the similarity between the cases can be determined, and the credibility and accuracy of the calculated similarity can be improved.

In an embodiment, the server calculates the target similarity between each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph; The preset coefficient and the similarity of each target are calculated, and the similarity between the first knowledge graph and each of the second knowledge graphs is calculated, that is, the unit of a second knowledge graph will belong to each of the second knowledge graphs. The target similarity is multiplied by the corresponding preset coefficient and then accumulated to obtain the similarity between the target case and the corresponding sentenced case. Wherein, each first entity node in the first knowledge graph corresponds to each second entity node in the second knowledge graph in a one-to-one correspondence. It should be noted that the aforementioned preset coefficients can be set based on actual conditions, and this application does not specifically limit this.

In an embodiment, the target similarity calculation method is specifically: calculating according to the node attribute information and/or node relationship information in the first knowledge graph and the node attribute information/or node relationship information in each second knowledge graph The target similarity between each first entity node and the corresponding second entity node in each second knowledge graph. Among them, the node attribute information is the attributes of the entity nodes in the knowledge graph, such as the specific parameters of the entity nodes such as dispute focus, fact elements, and evidence, and the node relationship information is the relationship information between the entity nodes in the knowledge graph, such as the dispute focus Relationship with factual elements.

In an embodiment, the method for calculating the target similarity is specifically: calculating each first entity node and each second entity node according to the node attribute information in the first knowledge graph and the node attribute information in each second knowledge graph. The first similarity between corresponding second entity nodes in the knowledge graph; and/or calculate each first entity node according to the node relationship information in the first knowledge graph and the node relationship information in each second knowledge graph The second degree of similarity with the corresponding second entity node in each of the second knowledge graphs; the first entity node between each first entity node and the corresponding second entity node in each second knowledge graph The similarity and/or the second similarity serve as the target similarity. Through the node attribute information, the semantic similarity between the cases can be obtained, and the logical similarity between the cases can be obtained through the node relationship information. The credibility and accuracy of similarity can be improved.

In an embodiment, when the target similarity is the first similarity and the second similarity, the calculation method of the similarity between the first knowledge graph and the second knowledge graph is specifically: multiplying each first similarity After accumulating the corresponding first weight coefficients, the first target similarity between the first knowledge graph and the second knowledge graph is obtained; each second similarity is multiplied by the corresponding second weight coefficient and then accumulated , The second target similarity between the first knowledge graph and the second knowledge graph is obtained; the sum of the first target similarity and the second target similarity will be calculated, and the difference between the first target similarity and the second target similarity will be calculated And as the similarity between the first knowledge graph and the second knowledge graph. Through node attribute information and node relationship information, the semantic and logical similarity between cases can be comprehensively considered, and the credibility and accuracy of similarity can be further improved.

It should be noted that the above-mentioned first weight coefficient and second weight coefficient can be set based on actual conditions, the first weight coefficient of each physical node can be the same or different, and the second weight coefficient of each physical node can be the same , Can also be different, and this application is not specifically limited.

Step S103: Determine a sentenced case similar to the target case according to the similarity between the target case and each of the sentenced cases.

The server determines the judged cases similar to the target case according to the similarity between the target case and each of the judged cases, that is, the judged case with the similarity greater than the preset similarity threshold is regarded as the judged case similar to the target case. It should be noted that the aforementioned similarity threshold may be set based on actual conditions, which is not specifically limited in this application.

Step S104: Determine the probability of victory of the target case based on a sentenced case similar to the target case.

The server determines the probability of victory of the target case based on the sentenced cases similar to the target case, that is, counts the number of first cases in which the defendant wins, the number of second cases in which the plaintiff wins, and the total number of cases in the sentenced cases similar to the target case, and Calculate the percentage of the number of first cases to the total number of cases, and calculate the percentage of the number of second cases to the total number of cases, and then calculate the percentage of the number of first cases to the total number of cases and/or the percentage of the number of second cases to the total number of cases As the probability of winning the target case.

In one embodiment, count the total number of cases that are similar to the target case and determine whether the total number of cases is greater than or equal to the preset number threshold; if the total number of cases is greater than or equal to the preset number threshold, then the statistics The number of first cases in which the defendant wins and the number of second cases in which the plaintiff wins in sentenced cases similar to the target case; calculate the percentage of the number of first cases to the total number of cases, and according to the percentage of the number of first cases to the total number of cases, Determine the probability of the defendant winning the target case; calculate the percentage of the number of second cases to the total number of cases, and determine the probability of the plaintiff winning the target case based on the percentage of the number of first cases to the total number of cases; compare the probability of winning the target case and/ Or the probability of the plaintiff winning the case is taken as the probability of winning the target case. If the total number of cases is less than the preset number threshold, a reminder message is sent to the terminal device.

Among them, the method for determining the probability of the defendant's victory or the probability of the plaintiff's victory may be specifically: taking the percentage of the number of first cases in the total number of cases as the probability of the defendant in the target case, or taking the number of second cases as the percentage of the total number of cases. The method for determining the probability of the defendant’s victory or the probability of the plaintiff’s victory can also be specifically: obtaining the pre-stored mapping table between the number of first cases or the number of second cases and the weight coefficient, and querying the mapping table to obtain the number of first cases or The weight coefficient corresponding to the number of second cases, and then calculate the product of the weight coefficient and the number of first cases or the percentage of the number of second cases to the total number of cases, and use this product as the probability of the defendant winning the target case or the probability of winning the plaintiff. It should be noted that the above-mentioned mapping relationship table between the number of prestored first cases or the number of second cases and the weight coefficient can be set based on actual conditions, and this application does not specifically limit this.

In one embodiment, the evidence type is determined according to the second knowledge graph corresponding to the judged case similar to the target case, and the determined evidence type is used as the target evidence type to be recommended, that is, the corresponding judged case with the greatest similarity is obtained And obtain the evidence knowledge information from the second knowledge graph, and then determine the evidence type to which the evidence knowledge information belongs, and use the determined evidence type as the target evidence type to be recommended. When receiving the case winning rate query request sent by the terminal device, the server obtains the winning probability and target evidence type of the corresponding case according to the case winning rate query request, and sends the winning probability and target evidence type to the terminal device.

The method for determining the case winning rate provided by the above embodiments calculates the similarity between the case to be predicted and each of the cases that have been judged based on the knowledge map of the case to be predicted for the probability of winning and the knowledge map of the case that has been judged. Predict the similarity between the case for predicting the probability of winning and each sentenced case, determine the sentenced case similar to the case, and then determine the probability of winning the case based on the sentenced case similar to the case, which can be fast and relatively accurate Determine the probability of winning the case.

Please refer to FIG. 3, which is a schematic flowchart of another method for determining a case winning rate provided by an embodiment of the application.

As shown in FIG. 3, the method for determining the case winning percentage includes steps S201 to 205.

Step S201: Obtain a first knowledge graph of a target case and a second knowledge graph of a plurality of judged cases, where the target case is a case for which the probability of winning the case is to be predicted.

Step S202: Check each second knowledge graph according to the first knowledge graph, and screen each second knowledge graph according to the verification result of each second knowledge graph .

Specifically, the first dispute focus of the target case is obtained from the first knowledge graph, and the second dispute focus of each sentenced case is obtained from each second knowledge graph; the category of the first dispute focus is correlated with each 2. Whether the category of the dispute focus is the same. If the category of the first dispute focus is the same as the category of the second dispute focus, it is determined that the corresponding second knowledge graph passes the verification, and if the category of the first dispute focus is the same as that of the second dispute focus If the categories are different, it is determined that the corresponding second knowledge graph fails the verification; the second knowledge graphs that fail the verification are eliminated, the second knowledge graphs that pass the verification are retained, and each second knowledge graph that passes the screening is obtained.

Step S203: Calculate the similarity between the target case and each of the sentenced cases corresponding to the first knowledge graph and each of the second knowledge graphs that have passed the screening.

The server calculates the similarity between the target case and each convicted case after the screening based on the first knowledge graph and the second knowledge graph of each convicted case after passing the screening. Specifically, the similarity between the first knowledge graph and each second knowledge graph that has passed the screening is calculated; the similarity between the first knowledge graph and each second knowledge graph that has passed the screening is taken as the target case and The similarity between each convicted case after screening.

Step S204: Determine a sentenced case similar to the target case according to the similarity between the target case and each of the sentenced cases.

Step S205: Determine the probability of victory of the target case based on the judged cases similar to the target case.

The method for determining the winning rate of a case provided in the above embodiment calculates the similarity between the case to be predicted for the probability of winning and each case based on the knowledge map of the case for which the probability of winning is to be predicted and the knowledge map of the judged case after being screened. , And based on the similarity between the case to be predicted and each sentenced case, determine a sentenced case similar to the case, and then determine the probability of victory of the case based on the sentenced case similar to the case. The knowledge graphs of judged cases are screened, which reduces the knowledge graphs involved in the calculation of similarity, which can increase the speed of calculation of similarity, thereby further improving the speed of determining the probability of winning the case.

Please refer to FIG. 4, which is a schematic block diagram of an apparatus for determining a case winning rate according to an embodiment of the application.

As shown in FIG. 4, the device 400 for determining the winning percentage of a case includes: an obtaining module 301, a calculating module 302, and a determining module 303.

The obtaining module 301 is configured to obtain a first knowledge graph of a target case and a second knowledge graph of multiple judged cases, where the target case is a case for which the probability of winning the case is to be predicted.

The calculation module 302 is configured to calculate the similarity between the target case and each of the sentenced cases according to the first knowledge graph and each of the second knowledge graphs.

The determining module 303 is configured to determine a sentenced case similar to the target case according to the similarity between the target case and each of the sentenced cases.

The determining module 303 is further configured to determine the probability of winning the target case based on a sentenced case similar to the target case.

In an embodiment, the calculation module 302 is further configured to calculate the similarity between the first knowledge graph and each of the second knowledge graphs; and compare the first knowledge graph with each of the first knowledge graphs. The similarity between the two knowledge graphs is taken as the similarity between the target case and each of the sentenced cases.

In one embodiment, the calculation module 302 is further configured to calculate the similarity between each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph. Degree; according to the respective preset coefficients corresponding to each of the first entity nodes and each of the target similarities, the similarity between the first knowledge graph and each of the second knowledge graphs is calculated.

In one embodiment, the calculation module 302 is further configured to calculate the node attribute information and/or node relationship information in the first knowledge graph and the node attribute information/or node in each of the second knowledge graphs. Relationship information, calculating the target similarity between each of the first entity nodes and the corresponding second entity nodes in each of the second knowledge graphs.

In one embodiment, the determining module 303 is also used to count the total number of cases sentenced similar to the target case, and to determine whether the total number of cases is greater than or equal to a preset number threshold; If the total number of cases is greater than or equal to the preset number threshold, the number of first cases in which the defendant wins and the number of second cases in which the plaintiff wins in the sentenced cases similar to the target case are counted; the number of first cases is calculated as a percentage The percentage of the total number of cases, and according to the percentage of the number of first cases to the total number of cases, determine the probability of the defendant winning the target case; calculate the percentage of the number of second cases to the total number of cases , And according to the percentage of the number of the first cases to the total number of cases, determine the probability of the plaintiff winning the target case; taking the probability of winning the defendant and/or the plaintiff of the target case as the winning probability of the target case Probability.

In one embodiment, the determining module 303 is further configured to determine the type of evidence according to the second knowledge graph corresponding to a sentenced case similar to the target case, and use the determined evidence type as the target to be recommended Type of evidence.

Please refer to FIG. 5, which is a schematic block diagram of another device for determining a case winning rate provided by an embodiment of the application.

As shown in FIG. 5, the device 400 for determining the winning percentage of a case includes: an acquisition module 401, a verification and screening module 402, a calculation module 403 and a determination module 404.

The obtaining module 401 is configured to obtain a first knowledge graph of a target case and a second knowledge graph of multiple judged cases, where the target case is a case for which the probability of winning the case is to be predicted.

The verification screening module 402 is configured to verify each of the second knowledge graphs according to the first knowledge graph, and perform verification on each of the second knowledge graphs according to the verification result of each second knowledge graph. 2. Screen the knowledge map.

The calculation module 403 is configured to calculate the similarity between the target case and each of the judged cases corresponding to the first knowledge graph and each of the second knowledge graphs that have passed the screening.

The determining module 404 is configured to determine a sentenced case similar to the target case according to the similarity between the target case and each of the sentenced cases.

The determining module 404 is configured to determine the probability of winning the target case based on a judged case similar to the target case.

It should be noted that those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working process of the above-described device and each module and unit can refer to the corresponding process in the embodiment of the method for determining the winning rate of a case. , I won’t repeat it here.

The apparatus provided in the foregoing embodiment may be implemented in the form of a computer program, and the computer program may run on the computer device as shown in FIG. 6.

Please refer to FIG. 6, which is a schematic block diagram of the structure of a computer device provided by an embodiment of the application. The computer device may be a server.

As shown in FIG. 6, the computer device includes a processor, a memory, and a network interface connected through a system bus, where the memory may include a non-volatile storage medium and an internal memory.

The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions. When the program instructions are executed, the processor can execute any method for determining the winning rate of a case.

The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any method for determining the winning rate of a case.

The network interface is used for network communication, such as sending assigned tasks. Those skilled in the art can understand that the structure shown in FIG. 6 is only a block diagram of part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may Including more or fewer parts than shown in the figure, or combining some parts, or having a different arrangement of parts.

It should be understood that the processor may be a central processing unit (Central Processing Unit, CPU), and the processor may also be other general-purpose processors or digital signal processors. (Digital Signal Processor, DSP), Application Specific Integrated Circuit (Application Specific Integrated Circuit, ASIC), field programmable gate array (Field-Programmable Gate Array, FPGA) Or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor.

Wherein, in one embodiment, the processor is used to run a computer program stored in a memory to implement the following steps: obtain a first knowledge graph of a target case and a second knowledge graph of a plurality of judged cases, where all The target case is the case for which the probability of winning the case is to be predicted; according to the first knowledge graph and each of the second knowledge graphs, the similarity between the target case and each of the judged cases is calculated; according to the The degree of similarity between the target case and each of the sentenced cases is determined to determine a sentenced case similar to the target case; and the probability of winning the target case is determined based on the sentenced case similar to the target case.

In one embodiment, when the processor realizes the calculation of the similarity between the target case and each of the sentenced cases according to the first knowledge graph and each of the second knowledge graphs, it uses To realize: calculate the similarity between the first knowledge graph and each of the second knowledge graphs; use the similarity between the first knowledge graph and each of the second knowledge graphs as the target The degree of similarity between the case and each of the said sentenced cases.

In an embodiment, each first entity node in the first knowledge graph corresponds to each second entity node in the second knowledge graph one-to-one, and the processor is implementing the calculation of the first knowledge graph When the similarity with each of the second knowledge graphs is used, it is used to realize: calculate each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph Target similarity between each of the first entity nodes; according to the respective preset coefficients corresponding to each of the first entity nodes and each of the target similarities, calculate the difference between the first knowledge graph and each of the second knowledge graphs Similarity.

In an embodiment, when the processor realizes the calculation of the target similarity between each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph, Used to realize: calculate each of the first entities according to the node attribute information and/or node relationship information in the first knowledge graph and the node attribute information/or node relationship information in each of the second knowledge graphs The target similarity between the node and the corresponding second entity node in each of the second knowledge graphs.

In one embodiment, when the processor is used to determine the probability of success of the target case based on a sentenced case similar to the target case, it is used to realize: statistics of the sentenced case similar to the target case The total number of cases, and determine whether the total number of cases is greater than or equal to the preset number threshold; if the total number of cases is greater than or equal to the preset number threshold, count defendants in sentenced cases similar to the target case The number of first cases won by the plaintiff and the number of second cases won by the plaintiff; the percentage of the number of first cases in the total number of cases is calculated, and the percentage of the number of first cases in the total number of cases is determined. The probability of the defendant winning the target case; calculating the percentage of the number of second cases to the total number of cases, and determining the probability of the plaintiff winning the target case based on the percentage of the number of first cases in the total number of cases ; The probability of winning the defendant and/or the probability of winning the plaintiff of the target case is taken as the probability of winning the target case.

In an embodiment, after the processor realizes that the probability of success of the target case is determined based on a sentenced case similar to the target case, it is also used to realize: according to a sentenced case similar to the target case Corresponding to the second knowledge graph, determine the evidence type, and use the determined evidence type as the target evidence type to be recommended.

Wherein, in another embodiment, the processor is configured to run a computer program stored in a memory to implement the following steps: obtain a first knowledge graph of a target case and a second knowledge graph of a plurality of judged cases, wherein, The target case is a case for which the probability of winning the case is to be predicted; according to the first knowledge graph, each of the second knowledge graphs is checked, and according to the verification result of each of the second knowledge graphs, each of the second knowledge graphs is checked. Screening by the second knowledge graphs; according to the first knowledge graphs and each of the second knowledge graphs that have passed the screening, calculate the similarity between the target case and each of the judged cases Degree; according to the similarity between the target case and each of the sentenced cases, determine the sentenced cases similar to the target case; determine the target case according to the sentenced cases similar to the target case The probability of winning.

It should be noted that those skilled in the art can clearly understand that for the convenience and brevity of the description, the specific working process of the computer device described above can refer to the corresponding process in the embodiment of the method for determining the winning rate of a case. Go into details again.

The embodiment of the present application also provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, the computer program includes program instructions, and the method implemented when the program instructions are executed can refer to this Various embodiments of the method for determining the winning percentage of the application case. Optionally, the computer-readable storage medium may be non-volatile or volatile.

The computer-readable storage medium may be the internal storage unit of the computer device described in the foregoing embodiment, for example, the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk or a smart memory card (Smart Memory Card) equipped on the computer device. Media Card, SMC), Secure Digital (SD) card, Flash Card, etc.

It should be understood that the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in the specification of this application and the appended claims, unless the context clearly indicates other circumstances, the singular forms "a", "an" and "the" are intended to include plural forms.

It should also be understood that the term "and/or" used in the specification of this application and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations. It should be noted that in this article, the terms "include", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements not only includes those elements, It also includes other elements that are not explicitly listed, or elements inherent to the process, method, article, or system. Without more restrictions, the element defined by the sentence "including a..." does not exclude the existence of other identical elements in the process, method, article, or system that includes the element.

The serial numbers of the foregoing embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments. The above are only specific implementations of this application, but the protection scope of this application is not limited to this. Anyone familiar with the technical field can easily think of various equivalents within the technical scope disclosed in this application. Modifications or replacements, these modifications or replacements shall be covered within the scope of protection of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

A method for determining the winning rate of a case, which includes:

Acquire a first knowledge graph of a target case and a second knowledge graph of multiple judged cases, where the target case is a case for which the probability of winning the case is to be predicted;

Calculating the similarity between the target case and each of the sentenced cases according to the first knowledge graph and each of the second knowledge graphs;

According to the similarity between the target case and each of the sentenced cases, determine the sentenced cases that are similar to the target case;

Determine the probability of success of the target case based on the sentenced cases similar to the target case.
The method for determining the winning rate of a case according to claim 1, wherein the calculation of the similarity between the target case and each of the judged cases is based on the first knowledge graph and each of the second knowledge graphs Degree, including:

Calculating the similarity between the first knowledge graph and each of the second knowledge graphs;

The similarity between the first knowledge graph and each of the second knowledge graphs is taken as the similarity between the target case and each of the judged cases.
The method for determining the winning rate of a case according to claim 2, wherein each first entity node in the first knowledge graph corresponds to each second entity node in the second knowledge graph one-to-one; said calculating said The similarity between the first knowledge graph and each of the second knowledge graphs includes:

Calculating the target similarity between each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph;

Calculate the similarity between the first knowledge graph and each of the second knowledge graphs according to the preset coefficient corresponding to each of the first entity nodes and each of the target similarities.
The method for determining the winning rate of a case according to claim 3, wherein said calculating the target between each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph Similarity, including:

According to the node attribute information and/or node relationship information in the first knowledge graph and the node attribute information/or node relationship information in each of the second knowledge graphs, calculate each of the first entity nodes and each The target similarity between corresponding second entity nodes in the second knowledge graph.
The method for determining the case winning rate according to any one of claims 1 to 4, wherein the determining the probability of winning the target case based on a sentenced case similar to the target case comprises:

Count the total number of judged cases similar to the target case, and determine whether the total number of cases is greater than or equal to a preset number threshold;

If the total number of cases is greater than or equal to the preset number threshold, count the number of first cases in which the defendant wins and the number of second cases in which the plaintiff wins among the sentenced cases similar to the target case;

Calculate the percentage of the number of the first cases to the total number of cases, and determine the probability of the defendant winning the target case according to the percentage of the number of the first cases to the total number of cases;

Calculating the percentage of the number of second cases in the total number of cases, and determining the probability of the plaintiff winning the target case according to the percentage of the number of first cases in the total number of cases;

The probability of winning of the defendant and/or the probability of winning of the plaintiff of the target case is taken as the probability of winning of the target case.
The method for determining the case winning rate according to any one of claims 1 to 4, wherein after determining the probability of winning of the target case based on a sentenced case similar to the target case, the method further comprises:

According to the second knowledge graph corresponding to the judged case similar to the target case, the evidence type is determined, and the determined evidence type is used as the target evidence type to be recommended.
The method for determining the success rate of a case according to any one of claims 1 to 4, wherein the calculation of the target case and each of the previous knowledge graphs is based on the first knowledge graph and each of the second knowledge graphs. The degree of similarity between judged cases, including:

Verify each of the second knowledge graphs according to the first knowledge graph, and screen each of the second knowledge graphs according to the verification result of each second knowledge graph;

According to the first knowledge graph and each of the second knowledge graphs that have passed the screening, the similarity between the target case and each of the sentenced cases is calculated.
A case winning rate determining device, wherein the case winning rate determining device includes:

The acquisition module is used to acquire a first knowledge graph of a target case and a second knowledge graph of multiple convicted cases, where the target case is a case for which the probability of winning the case is to be predicted;

A calculation module, configured to calculate the similarity between the target case and each of the sentenced cases according to the first knowledge graph and each of the second knowledge graphs;

The determining module is used to determine a sentenced case similar to the target case according to the similarity between the target case and each of the sentenced cases;

The determining module is further configured to determine the probability of winning the target case based on a sentenced case similar to the target case.
A computer device, wherein the computer device includes a processor, a memory, and a computer program stored on the memory and executable by the processor, wherein when the computer program is executed by the processor, the The following steps:

Acquire a first knowledge graph of a target case and a second knowledge graph of multiple judged cases, where the target case is a case for which the probability of winning the case is to be predicted;

Calculating the similarity between the target case and each of the sentenced cases according to the first knowledge graph and each of the second knowledge graphs;

According to the similarity between the target case and each of the sentenced cases, determine the sentenced cases that are similar to the target case;

Determine the probability of success of the target case based on the sentenced cases similar to the target case.
The computer device according to claim 9, wherein the calculation of the similarity between the target case and each of the judged cases is based on the first knowledge graph and each of the second knowledge graphs. , The concrete realization of the following steps:

Calculating the similarity between the first knowledge graph and each of the second knowledge graphs;

The similarity between the first knowledge graph and each of the second knowledge graphs is taken as the similarity between the target case and each of the judged cases.
The computer device according to claim 10, wherein each first entity node in the first knowledge graph corresponds to each second entity node in the second knowledge graph one-to-one; and the calculation of the first entity node For the similarity between the knowledge graph and each of the second knowledge graphs, the following steps are specifically implemented:

Calculating the target similarity between each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph;

Calculate the similarity between the first knowledge graph and each of the second knowledge graphs according to the preset coefficient corresponding to each of the first entity nodes and each of the target similarities.
The computer device according to claim 11, wherein said calculating the target similarity between each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph When, the specific steps are as follows:

According to the node attribute information and/or node relationship information in the first knowledge graph and the node attribute information/or node relationship information in each of the second knowledge graphs, calculate each of the first entity nodes and each The target similarity between corresponding second entity nodes in the second knowledge graph.
The computer device according to any one of claims 9 to 12, wherein the following steps are specifically implemented when determining the probability of success of the target case based on a sentenced case similar to the target case:

Count the total number of judged cases similar to the target case, and determine whether the total number of cases is greater than or equal to a preset number threshold;

If the total number of cases is greater than or equal to the preset number threshold, count the number of first cases in which the defendant wins and the number of second cases in which the plaintiff wins among the sentenced cases similar to the target case;

Calculate the percentage of the number of the first cases to the total number of cases, and determine the probability of the defendant winning the target case according to the percentage of the number of the first cases to the total number of cases;

Calculating the percentage of the number of second cases in the total number of cases, and determining the probability of the plaintiff winning the target case according to the percentage of the number of first cases in the total number of cases;

The probability of winning of the defendant and/or the probability of winning of the plaintiff of the target case is taken as the probability of winning of the target case.
The computer device according to any one of claims 9 to 12, wherein after determining the probability of success of the target case based on a sentenced case similar to the target case, the following steps are further implemented:

According to the second knowledge graph corresponding to the judged case similar to the target case, the evidence type is determined, and the determined evidence type is used as the target evidence type to be recommended.
A computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the following steps are implemented:

Acquire a first knowledge graph of a target case and a second knowledge graph of multiple judged cases, where the target case is a case for which the probability of winning the case is to be predicted;

Calculating the similarity between the target case and each of the sentenced cases according to the first knowledge graph and each of the second knowledge graphs;

According to the similarity between the target case and each of the sentenced cases, determine the sentenced cases that are similar to the target case;

Determine the probability of success of the target case based on the sentenced cases similar to the target case.
The computer-readable storage medium according to claim 15, wherein the calculation of the relationship between the target case and each of the judged cases is based on the first knowledge graph and each of the second knowledge graphs. For similarity, the following steps are specifically implemented:

Calculating the similarity between the first knowledge graph and each of the second knowledge graphs;

The similarity between the first knowledge graph and each of the second knowledge graphs is taken as the similarity between the target case and each of the judged cases.
The computer-readable storage medium according to claim 16, wherein each first entity node in the first knowledge graph corresponds to each second entity node in the second knowledge graph; When describing the similarity between the first knowledge graph and each of the second knowledge graphs, the following steps are specifically implemented:

Calculating the target similarity between each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph;

Calculate the similarity between the first knowledge graph and each of the second knowledge graphs according to the preset coefficient corresponding to each of the first entity nodes and each of the target similarities.
The computer-readable storage medium according to claim 17, wherein said calculating the difference between each first entity node in the first knowledge graph and the corresponding second entity node in each second knowledge graph When the target similarity is achieved, the following steps are specifically implemented:

According to the node attribute information and/or node relationship information in the first knowledge graph and the node attribute information/or node relationship information in each of the second knowledge graphs, calculate each of the first entity nodes and each The target similarity between corresponding second entity nodes in the second knowledge graph.
18. The computer-readable storage medium according to any one of claims 15 to 18, wherein the following steps are specifically implemented when determining the probability of success of the target case based on a sentenced case similar to the target case:

Count the total number of judged cases similar to the target case, and determine whether the total number of cases is greater than or equal to a preset number threshold;

If the total number of cases is greater than or equal to the preset number threshold, count the number of first cases in which the defendant wins and the number of second cases in which the plaintiff wins among the sentenced cases similar to the target case;

Calculate the percentage of the number of the first cases to the total number of cases, and determine the probability of the defendant winning the target case according to the percentage of the number of the first cases to the total number of cases;

Calculating the percentage of the number of second cases in the total number of cases, and determining the probability of the plaintiff winning the target case according to the percentage of the number of first cases in the total number of cases;

The probability of winning of the defendant and/or the probability of winning of the plaintiff of the target case is taken as the probability of winning of the target case.
The computer-readable storage medium according to any one of claims 15 to 18, wherein the following steps are further implemented after determining the probability of success of the target case based on a judged case similar to the target case:

According to the second knowledge graph corresponding to the judged case similar to the target case, the evidence type is determined, and the determined evidence type is used as the target evidence type to be recommended.