Recommendation Research And Application Of Legal Articles Based On Legal Judgment Documents

In recent years,with the rapid development of big data and artificial intelligence technology and intelligent assistance in the judicial field,the state has continued to promote judicial reform and research and proposed the construction of Smart Courts with smart judicial services as the core,and by accelerating the construction of Smart Courts,enhancing the degree of informatization in multiple links such as case acceptance and trial.Currently,in the judicial field,legal provisions set the rights and obligations of citizens and regulate the results of judicial trials,and it is a valuable basis for case trials.As the problem of "different judgments in the same case and inconsistent application of legal provisions" has gradually emerged,the study of an intelligently assisted method for recommending legal provisions and the application of integrated systems to improve the work efficiency of judges has great research significance and application value.The purpose of the legal recommendation is to automatically output the legal provisions applicable to the case after the text of the case is input through the machine.In the trial process of traditional cases,judges need to read laws and regulations to select appropriate legal provisions,and at the same time,they need to pay expensive lawyer consultation fees to understand the basic situation of the case and the possible outcome of the judgment.In response to the above problems,this article based on the judgment document data of judicial criminal cases,through the analysis of lots of criminal cases,comprehensively considers the co-occurrence relationship between the case description and the legal clauses and combines the characteristics of the legal clause recommendation task from the perspective of text classification.To start,study the method of recommending legal articles based on legal judgment documents,and finally realize the system of recommending and applying legal articles.The main work of the thesis is as follows.(1)Research on the rule recommendation algorithm based on legal judgment documents: This article uses three methods to experiment and analyze the rule recommendation algorithm,namely the rule recommendation algorithm based on the RCNN model,the rule recommendation algorithm based on the BERT-RCNN model,and Rule recommendation algorithm based on BERT-RCNN model and co-occurrence relationship.First,use the RCNN model as the baseline model to implement the rule recommendation algorithm;secondly,on this basis,introduce the BERT model to strengthen the ability to extract text features to improve the accuracy of the model;finally,use the co-occurrence relationship between the rules,Use the Text Rank algorithm to reorder the law articles to further improve the experimental effect.(2)Design and implementation of law recommendation system: Based on the research of law recommendation algorithm and according to the demand research and analysis,designing the law recommendation system with using WEB application framework Fast API and Vue.js framework to realize the establishment of law recommendation system.