7th International Conference on ICTCS 2022

Authors - Souraneel Mandal, Tanaya Das
Abstract - Natural Language Processing techniques can help in deriving insights from unstructured text. In Indian Judicial System, implementation of natural language processing based tools are found to be less in numbers. The majority of documents in Indian Judiciary are unstructured, and legal practitioners manually evaluate them to extract facts, information which requires time. Hence, natural language processing techniques like N-gram combined with some others text analytics approach can assist in analyzing a whole document to extract contextual meaning. Hence, in this paper the authors have proposed an N-gram analysis approach along with some text pre-processing technique and visualize extracted features from legal documents of Indian judicial system. The authors have primarily used techniques like count vector, tf-idf along with uni-gram and bi-grams and visualized them using various graphs i.e., heatmap and cluster-map. The graphs clearly visualize the significant parameters present in a legal document that help a legal professional to extract ad understand the important facts present inside a case. The authors have used some other text pre-processing techniques like tokenization, lemmatization, stop words removal, names removal, numerical value removal etc. Also, after pre-processing the authors TfIdftransform to extract the frequency and weighted-mean respectively. The most influential parameters serve the legal professionals in exploring the probable facts of the legal documents as mentioned in penal code sections of Indian Judiciary System.