Identifying Sentiment in Legal Case Judgments using Random Forest Classifier

This paper explores sentiment analysis in legal judgments using machine learning techniques. Six machine learning models, including Naïve Bayes, K Nearest Neighbors, Logistic Regression, Support Vector Machine, Random Forest, and Decision Tree, are used along with three data embeddings namely T5, RoBerta, and LegalBert. The study aims to evaluate the effectiveness of these models in sentiment analysis tasks tailored to legal documents. Each model is trained and evaluated based on the unique characteristics of legal texts. Hyperparameter tuning has also been performed for all the models. The focus is on achieving high accuracy, F1 score and understanding the interpretability of the models' predictions. The highest accuracy achieved was by Random Forest of 67.5%. The study aims to provide insights into the applicability of various machine learning algorithms and data embeddings for sentiment analysis in the legal domain.