An automated system to predict final judgment in Sri Lankan legal cases using NLP

Abstract

"In today's globalized world, the judiciary and legal system are expected to handle an increasing number of legal cases quickly and efficiently. However, the growing volume and complexity of legal cases present a challenge for judges and lawyers, which requires an improved functioning of the legal system. To address this problem, this research project proposes the use of machine learning to predict court decisions.

The technical solution involves the use of a transformer model to analyze legal documents and predict the final outcome of ongoing cases. The dataset used for this research project includes both Sinhala and English language legal cases. The transformer model has been trained on this dataset to accurately predict court decisions.

The test results of the model show an accuracy of 86%, indicating that the proposed solution has the potential to be an effective tool for the judicial and legal system. This research project demonstrates the value of using machine learning in the legal field and opens up new avenues for research in this area."