Method for automatically identifying law violation of civil case ending and execution

Abstract

The invention discloses a method for automatically identifying illegal execution at the end of a civil case. The method comprises the following specific operation steps: step 1, adding rules; step 2, rule analysis; step 3, document identification; and step 5, action processing. After activity execution law violation research and judgment are completed, data such as cases, legal bases, affirmative facts and case settlement modes corresponding to law violation situations can be automatically obtained for cases suspected to terminate the activity execution law violation, and research and judgment results are generated from the data in a visual form and displayed to case handling procurators. The method can be applied to terminating the execution of the judge, assisting the procurators in handling cases and making decisions, improving the case handling efficiency of the procurators, reducing the workload of manual examination, saving judicial resources, maintaining normal judicial order, improving judicial authority and judicial credibility, and protecting legitimate rights and interests of citizens, legal persons and other organizations in the field of civil judicial.

Description

Technical Field

[0001] The present invention relates to an automatic identification method, and in particular to a method for automatically identifying civil cases in which the execution is terminated in violation of the law, and is applicable to the civil prosecution field in the prosecution field. Background Art

[0002] The execution procedure is the final stage of the civil litigation procedure. It is an important link to ensure the smooth implementation of the effective judgment. It is directly related to whether the parties can realize their rights and interests in winning the case. Terminating this execution procedure means that the court temporarily terminates the execution procedure and closes the case for cases where there is no property available for execution, and resumes execution after the property is discovered. Its original intention is to set a technical closure standard to solve the problem of difficulty in closing execution cases. It means that the content determined by the effective judgment will not be effectively implemented for a long period of time, and the legal rights and interests of the applicant for execution will be difficult to achieve. The problem of terminating the irregularities in this execution procedure has always been a hot and difficult issue reflected by the people. This invention focuses on the problem of "difficult execution" and pays attention to the procedural and substantive standards of "termination" cases.

[0003] Termination of this execution procedure is closely related to the vital interests of the parties, and there are many irregularities in judicial practice. The court regards cases of "termination of this execution procedure within three months" as the focus of internal supervision. The procuratorate has made in-depth research and judgment on the operation of judicial power, and achieved the reduction of the number of cases and the focus on problematic cases by setting two levels of search rules. First, the case filing time and case closing time stated in the ruling to terminate this execution are extracted and calculated, and cases in which the execution procedure was terminated within three months are screened out. Second, by studying the procedural and substantive requirements of "termination", that is, searching for cases with a closing date of the end of the year, batch characteristics, and cases with a short time interval between filing and closing, the abuse of the "termination" procedure such as failure to exhaust property investigation and disposal measures is discovered and supervised; the information of the person to be executed is collided with the information of the dishonest person to be executed and the list of restricted high-consumption, and the illegality of dishonest punishment is discovered and supervised.

[0004] Due to the diversity of documents and other information in court files, the complexity of civil case analysis, and the large number of multi-source heterogeneous data, there is currently no intelligent method or system that can automatically identify and terminate the illegal execution of the above cases. Therefore, there is an urgent need to provide a method for automatically identifying and terminating the illegal execution of civil cases. Summary of the invention

[0005] In order to address the deficiencies of the above-mentioned technologies, the present invention provides a method for automatically identifying illegal termination of execution in civil cases. This method can be applied to civil termination of execution rulings in the civil judicial field, assist prosecutors in case handling decisions, improve prosecutors' case handling efficiency, reduce the workload of manual review, and thus save judicial resources, maintain normal judicial order, enhance judicial authority and judicial credibility, and protect the legitimate rights and interests of citizens, legal persons and other organizations.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for automatically identifying the illegality of terminating the execution of a civil case, which specifically includes the following steps:

[0007] Step S1, rule addition: configure the extracted attribute features on the data screening platform, then the rule engine calls the above attribute features through the API interface protocol, adds the termination of the current illegal judgment rule in the Easy Rules designer to form a judgment rule set, and then packages the added judgment rule set into an Easy Rules rule engine package;

[0008] Step 2 S2, rule analysis: configure the correspondence between the termination ruling and the Easy Rules rule engine package on the management page of the business system backend; automatically complete the rule analysis after obtaining the termination ruling information through the data exchange platform;

[0009] Step S3, text recognition: After obtaining the information of the termination order through the data exchange platform, the text recognition is automatically completed using natural language processing; the recognized plain text information is transmitted to the data screening platform, and the attribute features of the case are extracted according to the configured document attribute feature rules, and the extracted information is transmitted back to the data exchange platform for storage;

[0010] Step S4, action processing: The data exchange platform transmits the case attribute characteristics to the business system, and the business system automatically identifies the correspondence between the termination of the current execution violation and the Easy Rules rule engine package to complete the analysis and judgment rule call. When the conditions for terminating the current execution violation are met, the system automatically locates the location of the violation in the termination ruling and makes a visual annotation, which is then displayed to the prosecutor in charge of the case.

[0011] Preferably, in step S1, the rule adding process is completed by the rule adding module, and the attribute features of the identifiable cases that terminate the illegal and illegal situations of this execution are configured on the data screening platform, or regular expressions are written or directly assigned;

[0012] According to the analysis rules, in the Easy Rules designer, the attribute features are combined through a series of logical relational operators such as and, and, or, not, greater than, less than, equal to, not equal, greater than or equal to, and less than or equal to as the conditions after if, and actions are added according to then and otherwise to complete the rule configuration, which is then called by the rule engine through the interface protocol;

[0013] Among them, the rules include but are not limited to legal basis: citing corresponding laws and regulations; description of illegal circumstances: describing specific illegal circumstances; standards for determining facts: clarifying the conditions required for determining illegal facts; requirements for case closing methods: stipulating legal case closing methods.

[0014] Preferably, in step S2, the rule parsing process is completed by the rule parsing module, that is, the rule parsing module converts the rules into computer-executable logical judgment statements and stores them in the rule base, supporting dynamic update and expansion; by reading the attribute feature rules configured by the data screening platform and the judgment rule set of the rule designer, the preloading of the data required by the rule engine before the judgment is completed, and after the application server TongWeb on which the judgment rules for terminating the execution of the illegal situation are started, the rule xml data and attribute feature configuration are first loaded into the Context required for the judgment, and stored in the cache.

[0015] Preferably, in step S2, the rule parsing process includes the following steps:

[0016] S21: Read the attribute extraction rules configured by the rule adding module from the data screening platform, including the regular type and the assignment type, that is, complete the pre-loading of the attribute feature rules required to terminate the current execution of the illegal situation analysis rules for subsequent process calls;

[0017] S22: Read and parse the rule set data Easy Rules rule engine package completed by the Easy Rules rule designer. The rule set is composed of a common rule consisting of three parts: if, then, and otherwise, and a loop rule for loop iteration of a specified collection type object;

[0018] S23: The attribute feature data set configured in step S21 and the research and judgment rule decision set configured in step S22 are called and merged to form a set of rule formulas for terminating the current illegal judgment of the execution to be used, and the parameters in the formulas are the data of the termination ruling of the current illegal judgment case that is about to be terminated after program segmentation processing;

[0019] S24: repeat steps S21 to S23, enter a loop, and complete the preloading operation of the analysis and judgment rule formula corresponding to the illegal situation in the current execution in the cache;

[0020] S25: All the results configured in steps S21 to S24 are used to form a set of rule formulas for terminating the illegality assessment of this execution, so as to prepare for the subsequent submission of the case to terminate the execution ruling data.

[0021] Preferably, in S22, in the rule parsing code, it is necessary to read the content of the rule set Easy Rules rule engine package. The code needs to first obtain the constructed resource package information corresponding to the specified Easy Rules rule engine package ID through the KnowledgeService interface, and then create a specific KnowledgeSession object through the Easy Rules rule engine package, then insert the corresponding business object, and finally execute the specific rule call according to the message sent by the caller.

[0022] Preferably, in step S3, text recognition is the process of performing text recognition and analysis on document data in the form of Word or Excel, using natural language processing to perform word segmentation, part-of-speech tagging, entity recognition and semantic analysis on the document, extracting key information related to the violation, and terminating the execution of the violation by combining with the action processing module to locate the location of the violation and perform visual annotation.

[0023] Preferably, in step S3, the process of text recognition is as follows:

[0024] S31: Upload the word or Excel file of the ruling to terminate this execution to the designated directory of the judgment system. The file size is limited to 5M;

[0025] S32: judging whether the file is a word or Excel file according to the file extension, if the file extension is doc or docx, the current file is a word file, if the file extension is xlsx or xsl, the current file is an Excel file;

[0026] S33: Use the python docx library to extract all the text content in the document in segments;

[0027] S34: Use Python openxl library to extract all text contents in the document by Excel cell;

[0028] S35: Use natural language processing to extract data elements from the text by sample annotation, convert it into a json file, use the key information extraction UIE model + regular expression method to extract the required elements in the document, and store the extracted elements in the data screening platform;

[0029] S356: Manually backfill and correct the inaccurately identified content. This step is optional in the entire model for terminating the illegality assessment of this execution.

[0030] Preferably, in step S4, the action processing process is completed by the action processing module; after the illegal analysis rules for terminating this execution are integrated and processed by the background rule engine and the data of the ruling for terminating this execution, the illegal circumstances to be identified and the laws and regulations on which they are based are displayed to the prosecutor in a visual manner.

[0031] Preferably, in step S4, the action processing flow is specifically as follows:

[0032] S41: The rule engine for terminating the current execution violation combines the rule parsing module with the text recognition module to finally generate a data set of illegal situations for cases that have been terminated within three months since the filing of the execution case, and the ModelAndView object obtains the illegal situation data set through ZbModel to form JSON data;

[0033] S42: Use the JSON data obtained by the ModelAndView object to push detailed illegal situations to the JSP page. The auxiliary case handling information pushed includes laws and regulations, facts, recommended case closing methods, and the names of the files and documents corresponding to the illegal situations, and the location data marked by the visual logo, to assist the prosecutor in analyzing the termination of the illegal execution case;

[0034] S43: The JSP page receives the data from the ModelAndView object, disassembles the attributes in the object, including laws and regulations, determined facts, and recommended case closing methods, and displays them to the prosecutor at the corresponding position of the function page, assisting the prosecutor in analyzing and judging the case of terminating the illegal execution and finally completing the case handling.

[0035] Preferably, in S41, the illegal situation data set includes several illegal situations triggered in the case, as well as the laws and regulations corresponding to the illegal situations, the illegal facts identified in the case, and the case closing method recommended to the prosecutor for such illegal situations.

[0036] The present invention discloses a method for automatically identifying the termination of the execution violation in a case, which can be applied in the field of civil justice to identify the termination of the execution violation procedure in a civil case. It extracts the attribute characteristics of the termination of the execution ruling data of the civil case, constructs the research and judgment rules, and intelligently identifies the illegal circumstances involved in the termination of the execution violation in the case. When the termination of the execution violation conditions is met, the illegal location is automatically located and the specific illegal circumstances, legal basis, etc. are visually marked to assist prosecutors in making decisions, thereby improving the prosecutors' case handling efficiency, saving judicial resources, maintaining the order of trial execution, and enhancing judicial authority and judicial credibility. At the same time, the automatic identification of the termination of the execution violation has become a reality, which has a very good development prospect. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a diagram showing the overall architecture of the method for automatically identifying cases and terminating illegal execution in the present invention.

[0038] Figure 2 This is a flow chart of the rule parsing phase of step 2 of the present invention.

[0039] Figure 3 This is a flow chart of the character recognition step of step three of the present invention.

[0040] Figure 4 This is a flow chart of the action processing link of step 4 of the present invention. DETAILED DESCRIPTION

[0041] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0042] like Figure 1 The method shown in the figure is to automatically identify civil cases and terminate the execution violation, which includes four steps: rule addition, rule analysis, text recognition, and action processing. After completing the illegal judgment of the execution activity, for the case suspected of illegal execution activities, the location, legal basis, facts, case closing method and other data corresponding to the illegal situation can be automatically obtained. These data will be generated in a visual form to show the judgment results to the prosecutor handling the case. The specific process is as follows:

[0043] Step 1: Add rules: Configure the extracted attribute features on the data screening platform, and then the rule engine calls these attribute features through the API interface protocol. Add the rules for terminating the current illegal execution to form a judgment rule set on the rule design page. The rules include legal basis, description of illegal circumstances, criteria for determining facts, and case closing methods. Then package the added judgment rule set into the Easy Rules rule engine package.

[0044] Among them, the data screening platform, rule engine, and rule designer can all adopt commonly used technologies in the industry. The data screening platform is a data distribution center, which can send the data of different projects (or business systems) to the target system after format packaging according to the requirements. The platform integrates data receiving components, data format packaging components, and data sending components. Through the configuration and combination of data pipelines, it realizes the flexibility of the overall process from data reception to storage, and ensures asynchronous processing of data to improve data processing performance. The rule engine is developed from the reasoning engine. It is a component embedded in the application program, which realizes the separation of business decisions from application code, and uses predefined semantic modules to write business decisions, accept data input, interpret business rules, and make business decisions according to business rules. The extracted attribute features are the structural element keywords extracted from the identification of cases or file documents. The recognition rules are written manually and automatically extracted by computer programs.

[0045] The rule adding process is completed by the rule adding module, which configures the attribute characteristics of the identifiable cases that terminate the illegal and illegal situations in the data screening platform, or writes regular expressions or directly assigns values. The rules include but are not limited to legal basis: citing relevant laws and regulations; description of illegal situations: describing specific illegal situations; criteria for determining facts: clarifying the conditions required for determining illegal facts; requirements for closing cases: stipulating legal closing methods.

[0046] Next, according to the analysis rules, in the Easy Rules designer, the attribute features are combined as the conditions after "if" through a series of logical relationship operators such as and, or, not, "greater than", "less than", "equal to", "not equal to", "greater than or equal to", and "less than or equal to", and actions are added according to "then" and "otherwise", thus completing the rule configuration and providing the rule engine with a call through the interface protocol. The configured rules can be modified and deleted.

[0047] Step 2: Rule parsing: Configure the correspondence between the ruling to terminate this execution and the EasyRules rule engine package on the management page of the business system backend; automatically complete the rule parsing after obtaining the information of the ruling to terminate this execution through the data exchange platform;

[0048] Among them, in the management background of the business system, the Easy Rules rule engine package can be set up to correspond with the execution ruling and the final case. Since the rule designer has designed the analysis and judgment rule set to be packaged into the Easy Rules rule engine package, a correspondence between the case cause and the analysis and judgment rule set is formed. For the data exchange platform, the commonly used data exchange platform in the industry is used. It is the entry platform for terminating this execution ruling. The structured elements of the termination of this execution ruling can be extracted, and the extracted structured elements are sent to the rule engine for analysis. The data exchange platform also has an interface with text recognition, which can recognize the text in the file.

[0049] The rule parsing process is completed by the rule parsing module, which converts the rules into computer-executable logical judgment statements and stores them in the rule base, supporting dynamic updates and expansions. By reading the attribute feature rules configured by the data screening platform and the judgment rule set (also called the judgment rule decision set) designed by the rule designer, the rule engine completes the pre-loading of data required before the judgment. After the application server TongWeb that the judgment rules for terminating the execution of illegal situations are started, the rule xml data and attribute feature configuration are first loaded into the Context required for the judgment and stored in the cache.

[0050] Specifically: Automatically identifying the illegal and unlawful situations in the case that require the termination of the current execution requires a series of operations to be completed through the rule parsing module, so that the rule engine can preload the required data before the judgment. The preloading method can also effectively improve the operating efficiency of the judgment.

[0051] After the application server TongWeb that the judgment rules for terminating the illegal execution of this time is started, the rule XML data and attribute feature configuration must first be loaded into the Context required for the judgment and stored in the cache to improve the calling efficiency, such as Figure 2 As shown in the figure, the rule parsing process is divided into the following stages:

[0052] S001: Read the attribute extraction rules configured by the rule adding module from the data screening platform. The attribute extraction rules here will be used in the analysis and judgment rules, including regular types and assignment types. For example, one of the analysis and judgment rules for the illegal situation of "less than three months from the filing of the execution to the termination of the execution procedure" to terminate the current execution violation is: through structured analysis of legal documents and extraction and calculation of the execution filing and closing dates, the execution filing date and closing date here are the necessary attribute features required for the analysis and judgment rules to terminate the current execution violation. Step S001 completes the preloading of the attribute feature rules required for the analysis and judgment rules to terminate the current execution violation for subsequent process calls.

[0053] S002: Read and parse the rule set data Easy Rules rule engine package completed by the Easy Rules rule designer. The rule set is also called a decision set. The rule set consists of ordinary rules consisting of three parts: if, then, and otherwise, and loop rules that can specify collection type objects for loop iteration.

[0054] In the rule parsing code, it is necessary to read the content of the rule set Easy Rules rule engine package. The code must first obtain the constructed resource package information corresponding to the specified Easy Rules rule engine package ID through the KnowledgeService interface, and then create a specific KnowledgeSession object through the Easy Rules rule engine package. Next, insert the relevant business objects, and finally execute the specific rule call according to the message sent by the caller. Step S002 mainly completes the loading of the rule set data.

[0055] S003: The attribute feature data set configured in step S001 is called and integrated with the research and judgment rule decision set configured in step S002 to form a set of formulas for terminating the illegal judgment of this execution. The parameters in the formula are the data of the termination ruling of this execution that will be terminated after program segmentation processing.

[0056] S004: Repeat steps S001 to S003 and enter a loop until the configuration of the rule adding phase is completed and the analysis and judgment rule formulas corresponding to the illegal situations of this execution are preloaded in the cache.

[0057] S005: This step forms a set of rule formulas for terminating the current execution violation assessment with all the results configured from S001 to S004, preparing for the subsequent submission of the case termination ruling data.

[0058] Step 3: Text recognition: After obtaining the information of the termination order through the data exchange platform, the text recognition is automatically completed using natural language processing technology; the recognized plain text information is transmitted to the data screening platform, and the attribute features of the case are extracted according to the configured document attribute feature rules, and the extracted information is transmitted back to the data exchange platform for storage;

[0059] Text recognition is the process of identifying and analyzing text. It uses natural language processing technology to segment documents, tag parts of speech, identify entities, and perform semantic analysis to extract key information related to illegal situations. In other words, text recognition is the process of intelligently identifying and analyzing document data in word or Excel format. Figure 3 As shown, the details are as follows:

[0060] L001: Upload the word or Excel file of the ruling to terminate this execution to the designated directory of the judgment system. The file size limit is within 5M;

[0061] L002: Determine whether it is a word or Excel file based on the file extension. If the extension is doc or docx, the current file is a word file. If the extension is xlsx or xsl, it is an Excel file.

[0062] L003: Use the python docx library to extract all the text content in the document in sections;

[0063] L004: Use Python openxl library to extract all text contents in documents by Excel cell;

[0064] L005: Use natural language processing technology to extract data elements from the text by annotating samples. For example, annotate 20 rulings on the termination of this execution, including the name of the court, case number, party information, agent information, litigation request, trial findings, court opinion, and judgment results. Convert them into json files, use the key information extraction UIE model + regular expression method to extract the required elements in the document, and store the extracted elements in the data screening platform.

[0065] L006: Due to certain defects in text recognition accuracy, it is necessary to manually fill in and correct the inaccurately recognized content. This step is optional in the entire model for terminating the illegal execution. This design can still run automatically when the amount of file data is large.

[0066] Step 4. Action processing: The data exchange platform transmits the case attribute characteristics to the business system, which automatically identifies the corresponding relationship between the termination of the execution violation and the data screening platform to complete the analysis and judgment rule call. When the conditions for terminating the execution violation are met, the system automatically locates the illegal situation in the termination ruling and visually marks it, and displays it to the prosecutor in charge.

[0067] The action processing process is completed through the action processing module; the illegal judgment rules for terminating this execution are integrated and processed by the background rule engine and the data of the termination order, and the illegal situations to be identified and the laws and regulations based on them are displayed to the prosecutor in a visual way; the illegal judgment rules engine for terminating this execution combines the rule parsing module and the text recognition module to finally generate a data set of illegal situations for cases that have been filed for less than three months but have been terminated. The illegal situation data set includes several illegal situations that were triggered in the case. The action processing workflow is as follows: Figure 4 As shown, the details are as follows:

[0068] A001: The rule engine for terminating this execution of illegality analysis and judgment, combined with the rule parsing module and the text recognition module, finally generates a data set of illegal situations of "private lending, recovery of labor remuneration". The illegal situation data set includes several illegal situations triggered in the case, as well as the laws and regulations corresponding to the illegal situations, the illegal facts identified in the case, and the recommended closing method for the prosecutor for such illegal situations.

[0069] A002: The rule controller is mainly responsible for parsing user input and processing background data to return corresponding results in the product application framework. It plays a role in connecting the front and back ends in the application. It can process the front-end data in the controller and then pass it to the back-end Service, and it can also pass the data processed by the back-end Service to the front-end page through the controller. The ModelAndView object is used to push detailed violations to the JSP page. In addition, it also includes laws and regulations, facts, recommended closing methods, and the names of the files and documents corresponding to the violations, and the location data marked by the visual logo.

[0070] A003: The JSP page receives the data from the ModelAndView object, disassembles the attributes in the object, including laws and regulations, determined facts, and recommended case closing methods, and displays them to the prosecutor at the corresponding location on the function page, assisting the prosecutor in analyzing and judging the case of terminating the illegal execution and ultimately completing the case handling.

[0071] The method disclosed in the present invention is a method for automatically identifying the termination of illegal execution in a case, which can automatically identify the file information, extract its attribute characteristics, and design rules based on the Easy Rules rule engine designer (URule is a pure Java rule engine, which is based on the RETE algorithm and provides six types of rule definition methods including wizard rule sets, script rule sets, decision tables, cross decision tables, decision trees, scoring cards and rule flows) to assist in completing intelligent analysis and judgment. When the conditions for terminating the illegal execution are met, the illegal position is automatically located in the file and the specific illegal circumstances, legal basis, etc. are visually identified to assist prosecutors in making decisions, thereby improving the efficiency of prosecutors in handling cases.

[0072] The present invention has the following beneficial effects:

[0073] (1) Developing a unique and quantifiable set of identification rules to end the current illegal execution can greatly improve the efficiency of prosecutors.

[0074] (2) Introducing text recognition, image recognition, semantic recognition, and deep learning technologies in artificial intelligence, and applying rule engine technology to the civil prosecution supervision industry, will make the automatic identification of illegal and unlawful situations in this execution a reality.

[0075] (3) Solved the problem of identifying and extracting business-specific vocabulary from text and the accuracy. Specific method: When doing Chinese word segmentation, define business named entity recognition, convert the original data into word annotations according to the BMEO rules, and use the Bilstm+crf model for model training. In order to improve the accuracy of model training, the entities extracted by the model are intersected with all entities in the data set, and the intersection and entities are calculated using accuracy, recall and f-value.

[0076] (4) It saves judicial resources, maintains the order of trial and execution, and enhances judicial authority and judicial credibility.

[0077] The above implementation modes are not limitations of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the scope of the technical solution of the present invention also belong to the protection scope of the present invention.

Claims

1. A method for automatically identifying civil cases and terminating the current execution violation, characterized by: The specific steps include: Step S1, rule addition: configure the extracted attribute features on the data screening platform, then the rule engine calls the above attribute features through the API interface protocol, adds the termination of the current illegal judgment rule in the Easy Rules designer to form a judgment rule set, and then packages the added judgment rule set into an Easy Rules rule engine package; Step 2 S2, rule analysis: configure the correspondence between the termination of this execution ruling and the EasyRules rule engine package on the management page of the business system backend; After obtaining the information of the ruling to terminate the execution through the data exchange platform, the rule parsing is automatically completed; Step S3, text recognition: After obtaining the information of the termination order through the data exchange platform, the text recognition is automatically completed using natural language processing; the recognized plain text information is transmitted to the data screening platform, and the attribute features of the case are extracted according to the configured document attribute feature rules, and the extracted information is transmitted back to the data exchange platform for storage; Step S4, action processing: The data exchange platform transmits the case attribute characteristics to the business system, and the business system automatically identifies the correspondence between the termination of the current execution violation and the Easy Rules rule engine package to complete the analysis and judgment rule call. When the conditions for terminating the current execution violation are met, the system automatically locates the location of the violation in the termination ruling and makes a visual annotation, which is then displayed to the prosecutor in charge of the case.

2. The method for automatically identifying civil cases and terminating the current execution violation according to claim 1 is characterized by: In the step S1, the rule adding process is completed by the rule adding module, and the attribute features of the identifiable cases that terminate the illegal and illegal situations of this execution are configured on the data screening platform, or regular expressions are written or directly assigned; According to the analysis rules, in the Easy Rules designer, the attribute features are combined through a series of logical relational operators such as and, and, or, not, greater than, less than, equal to, not equal, greater than or equal to, and less than or equal to as the conditions after if, and actions are added according to then and otherwise to complete the rule configuration, which is then called by the rule engine through the interface protocol; Among them, the rules include but are not limited to legal basis: citing corresponding laws and regulations; description of illegal circumstances: describing specific illegal circumstances; standards for determining facts: clarifying the conditions required for determining illegal facts; requirements for case closing methods: stipulating legal case closing methods.

3. The method for automatically identifying civil cases and terminating the current execution violation according to claim 1 is characterized by: In the step S2, the rule parsing process is completed by the rule parsing module, that is, the rule parsing module converts the rules into computer-executable logical judgment statements and stores them in the rule library, supporting dynamic update and expansion; by reading the attribute feature rules configured by the data screening platform and the judgment rule set of the rule designer, the rule engine completes the preloading of data required before the judgment, and after the application server TongWeb on which the judgment rules for terminating the execution of the illegal situation are started, the rule xml data and attribute feature configuration are first loaded into the Context required for the judgment and stored in the cache.

4. The method for automatically identifying civil cases and terminating the current execution violation according to claim 3 is characterized by: In step S2, the rule parsing process includes the following steps: S21: Read the attribute extraction rules configured by the rule adding module from the data screening platform, including the regular type and the assignment type, that is, complete the pre-loading of the attribute feature rules required to terminate the current execution of the illegal situation analysis rules for subsequent process calls; S22: Read and parse the rule set data Easy Rules rule engine package completed by the Easy Rules rule designer. The rule set is composed of a common rule consisting of three parts: if, then, and otherwise, and a loop rule for loop iteration of a specified collection type object; S23: The attribute feature data set configured in step S21 and the research and judgment rule decision set configured in step S22 are called and merged to form a set of rule formulas for terminating the current illegal judgment of the execution to be used, and the parameters in the formulas are the data of the termination ruling of the current illegal judgment case that is about to be terminated after program segmentation processing; S24: repeat steps S21 to S23, enter a loop, and complete the preloading operation of the analysis and judgment rule formula corresponding to the illegal situation in the current execution in the cache; S25: All the results configured in steps S21 to S24 are used to form a set of rule formulas for terminating the illegality assessment of this execution, so as to prepare for the subsequent submission of the case to terminate the execution ruling data.

5. The method for automatically identifying civil cases and terminating the current execution violation according to claim 1 is characterized by: In S22, in the rule parsing code, it is necessary to read the content of the rule set Easy Rules rule engine package. The code needs to first obtain the constructed resource package information corresponding to the specified Easy Rules rule engine package ID through the KnowledgeService interface, and then create a specific KnowledgeSession object through the Easy Rules rule engine package, then insert the corresponding business object, and finally execute the specific rule call according to the message sent by the caller.

6. The method for automatically identifying civil cases and terminating the current execution violation according to claim 1 is characterized by: In step S3, text recognition is the process of text recognition and analysis of document data in the form of Word or Excel, using natural language processing to perform word segmentation, part-of-speech tagging, entity recognition and semantic analysis on the document, extracting key information related to the violation, and terminating the execution of the violation by combining with the action processing module to locate the location of the violation and perform visual annotation.

7. The method for automatically identifying civil cases and terminating the current execution violation according to claim 6 is characterized by: In step S3, the process of text recognition is as follows: S31: Upload the word or Excel file of the ruling to terminate this execution to the designated directory of the judgment system. The file size is limited to 5M; S32: judging whether the file is a word or Excel file according to the file extension, if the file extension is doc or docx, the current file is a word file, if the file extension is xlsx or xsl, the current file is an Excel file; S33: Use the python docx library to extract all the text content in the document in segments; S34: Use Python openxl library to extract all text contents in the document by Excel cell; S35: Use natural language processing to extract data elements from the text by sample annotation, convert it into a json file, use the key information extraction UIE model + regular expression method to extract the required elements in the document, and store the extracted elements in the data screening platform; S356: Manually backfill and correct the inaccurately identified content. This step is optional in the entire model for terminating the illegality assessment of this execution.

8. The method for automatically identifying civil cases and terminating the current execution violation according to claim 1 is characterized by: In step S4, the action processing process is completed through the action processing module; after the illegal analysis rules for terminating this execution are integrated and processed by the background rule engine and the data of the ruling for terminating this execution, the illegal circumstances to be identified and the laws and regulations on which they are based are displayed to the prosecutor in a visual manner.

9. The method for automatically identifying civil cases and terminating the current execution violation according to claim 8 is characterized by: In step S4, the action processing flow is as follows: S41: The rule engine for terminating the current execution violation combines the rule parsing module with the text recognition module to finally generate a data set of illegal situations for cases that have been terminated within three months since the filing of the execution case, and the ModelAndView object obtains the illegal situation data set through ZbModel to form JSON data; S42: Use the JSON data obtained by the ModelAndView object to push detailed illegal situations to the JSP page. The auxiliary case handling information pushed includes laws and regulations, facts, recommended case closing methods, and the names of the files and documents corresponding to the illegal situations, and the location data marked by the visual logo, to assist the prosecutor in analyzing the termination of the illegal execution case; S43: The JSP page receives the data from the ModelAndView object, disassembles the attributes in the object, including laws and regulations, determined facts, and recommended case closing methods, and displays them to the prosecutor at the corresponding position of the function page, assisting the prosecutor in analyzing and judging the case of terminating the illegal execution and finally completing the case handling.

10. The method for automatically identifying civil cases and terminating the current execution violation according to claim 9 is characterized by: In S41, the illegal situation data set includes several illegal situations triggered in the case, as well as the laws and regulations corresponding to the illegal situations, the illegal facts identified in the case, and the case closing method recommended to the prosecutor for such illegal situations.

Images