Iterative concept-based clustering of Indian court judgments

Abstract

This paper proposes architecture for the legal practitioners to lessen the burden of reading full document. The iterative Latent Semantic Analysis (LSA)-based concept extracting and clustering of legal judgment is proposed here. Headnotes in legal judgment explain about the cases in few sentences which is not enough for the legal practitioners for preparing the arguments. Our approach is to automate text processing for main concepts retrieval of legal judgments. An iterative latent semantic-based concept extraction is used. The Natural Language Processing (NLP) and data mining techniques are used for the comparison of full judgment against: iterative latent semantic analysis based concept retrieval, headnote, and different summarization levels using the existing summarization tool.