Legal Cause Prediction With Inner Descriptions And Outer Hierarchies

Legal Cause Prediction (LCP) aims to determine the charges in criminal cases or types of disputes in civil cases according to the fact descriptions. The research to date takes LCP as a text classification task and fails to consider the outer hierarchical dependencies and inner text information of causes. However, this information is critical for understanding causes and is expected to benefit LCP. To address this issue, we propose the Hierarchical Legal Cause Prediction (HLCP) model to incorporate this crucial information within the seq2seq framework. Specifically, we employ an attention-based seq2seq model to predict the cause path and utilize the inner text information to filter out noisy information in fact descriptions. We conduct experiments on 4 real-world criminal and civil datasets. Experimental results show that our model achieves significant and consistent improvements over all baselines.