ILDC for CJPE: Indian Legal Documents Corpus for Court Judgment Prediction and Explanation

An automated system that could assist a judge in predicting the outcome of a\ncase would help expedite the judicial process. For such a system to be\npractically useful, predictions by the system should be explainable. To promote\nresearch in developing such a system, we introduce ILDC (Indian Legal Documents\nCorpus). ILDC is a large corpus of 35k Indian Supreme Court cases annotated\nwith original court decisions. A portion of the corpus (a separate test set) is\nannotated with gold standard explanations by legal experts. Based on ILDC, we\npropose the task of Court Judgment Prediction and Explanation (CJPE). The task\nrequires an automated system to predict an explainable outcome of a case. We\nexperiment with a battery of baseline models for case predictions and propose a\nhierarchical occlusion based model for explainability. Our best prediction\nmodel has an accuracy of 78% versus 94% for human legal experts, pointing\ntowards the complexity of the prediction task. The analysis of explanations by\nthe proposed algorithm reveals a significant difference in the point of view of\nthe algorithm and legal experts for explaining the judgments, pointing towards\nscope for future research.\n