Overview of CAIL2018: Legal Judgment Prediction Competition

TL;DR

This paper surveys the CAIL2018 Legal Judgment Prediction competition, outlining three prediction tasks built on a large-scale Chinese criminal corpus. It highlights the shift to neural models for law article and charge prediction while noting the relative difficulty of predicting prison terms and the value of tricks like domain-specific embeddings, data balancing, joint learning, and feature engineering. The contribution includes releasing the dataset and synthesizing effective modeling strategies, offering practical guidance for future legal NLP research. Overall, CAIL2018 sets a strong benchmark and accelerates progress in NLP-driven legal judgment prediction.

Abstract

In this paper, we give an overview of the Legal Judgment Prediction (LJP) competition at Chinese AI and Law challenge (CAIL2018). This competition focuses on LJP which aims to predict the judgment results according to the given facts. Specifically, in CAIL2018 , we proposed three subtasks of LJP for the contestants, i.e., predicting relevant law articles, charges and prison terms given the fact descriptions. CAIL2018 has attracted several hundreds participants (601 teams, 1, 144 contestants from 269 organizations). In this paper, we provide a detailed overview of the task definition, related works, outstanding methods and competition results in CAIL2018.