Comparative Analysis of Legal Outcome Prediction with Detailed and Summarized Text

Legal Outcome Prediction(LOP) is the process of predicting the possible outcome of a court case based on the contents in the case file.The aim of this work is to perform a comparative analysis to assess the effectiveness of popular language models for outcome prediction with the case file as well as its summary content.Pretrained models of BERT, LegalBERT, DistilBERT and RoBERTa are enhanced with transfer learning on the legal data files and the models are evaluated with accuracy and F-score. It is observed that RoBERTa performs best in LOP for both complete case file text as well as for summary text.