Research And Implementation Of Case Plot Identification System Based On Semantic Equivalence

Combining AI technology with the judiciary is a key step in modernizing the judiciary.There are many problems in the current case handling process when completing the case plot identification,such as many case files to be processed but few staff,serious oral description,many documents involved,difficulty in inferring the circumstances,and great influence on sentencing.How to extract key information from these data and make semantic equivalence judgment has become the focus of research.The key steps of case plot identification include case plot extraction and case plot semantic equivalence determination.For case plot extraction,this paper proposes a case plot extraction model based on extended TextCNN.This model is based on the dossier data characteristics of case plot identification in the judicial field and the limitations of the TextCNN model in dealing with tasks in this field.By randomly removing dependent features,adding irrelevant data sets and setting loss weights for multiple training to adjust the bias,the text classification effect of case plot extraction has been improved to a certain extent.For the semantic equivalence judgment of case plot,this paper proposes a case plot semantic equivalence judgment model based on WWM-BERT.The multi-head attention mechanism in the improved model can comprehensively obtain the contextual semantics of the plot behavior of the case.The pre-trained model can alleviate the problem of data dependence and is more suitable for judicial application scenarios.The whole-word masking method is suitable for the scenario of semantic similarity judgment of Chinese short texts.In addition,the LTP word segmentation tool that integrates machine learning algorithms is also used.Using the case plot extraction model and the case plot semantic equivalence model proposed in this paper,a case plot identification system is developed.This case plot identification system includes modeling functions and business functions.This article focuses on business functions.Perform requirements analysis more intuitively by drawing use case diagrams.In the part of system design and implementation,it includes frame design,interface design,database design and main function design.Design the system and describe the implementation process by using tables,drawing class diagrams,and functional sequence diagrams.Develop a case plot identification system based on the background of smart judicial construction.