Legal Judgment Prediction with Multiple Perspectives on Civil Cases
Lili Zhao,Linan Yue,Yanqing An,Ye Liu,Kai Zhang,Weidong He,Yanmin Chen,Senchao Yuan,Qi Liu
DOI: https://doi.org/10.1007/978-3-030-93046-2_60
IF: 14.4
2021-01-01
Artificial Intelligence
Abstract:Legal Judgment Prediction, which aims at predicting the judgment result based on case materials, is an essential task in Legal Intelligence. Most existing studies have analyzed and modeled criminal cases as a whole, while a small part of the research focuses on civil cases which are also significant in the legal system. In these studies, most model on a certain type of civil causes and make the judgment prediction based on the ascertained fact from the perspective of the court. However, in real-world scenarios, there are different civil cases on various causes, and every case is judged from multiple perspectives (i.e., plaintiff, defendant, and the court). It is difficult to make judgment predictions on various civil causes. To address the above challenges, in this paper, we propose a novel Civil Case Judgment (CCJudge) prediction method by simulating the logic of judges. Following this logic, we construct an external knowledge base that contains the explanations and applications of every cause, it helps make the judgment on various causes. Furthermore, a special encoder layer and interaction layer are designed for learning linguistic semantics from multiple perspectives. We conduct extensive experiments on a real-world dataset. The experimental results demonstrate the effectiveness of our method.
computer science, artificial intelligence