A Neural Network Based Method for Judging the Rationality of Litigation Request

摘要

With the advancement of the construction of the intelligence court, artificial intelligence technology has become more and more significant in the litigation risk analysis. The reasonable prediction of the litigation request can guide the parties to form a reasonable litigation expectation and guide the case diversion; it can also provide intelligent risk warning to assist judges. The current litigation request reasonable prediction method has a low accuracy rate, and lacks a method for rationalizing the expectation of heterogeneous and diverse cases; at the same time, based on machine learning methods, some research focus on the rationality judgment of litigation requests. However, the rationality of litigation request feature design, selection and model are still to be explored. In this paper, we propose a rationality prediction framework for litigation requests based on multi-party evidence correlation model and neural network. Firstly, based on the judicial knowledge of litigation request and the multi-party evidence association model, the judicial characteristics of the reasonableness of litigation request is designed. Then we train and predict the judicial features based on the deep neural network model. Finally, the prediction and evaluation of the reasonableness of litigation is realized.