Relevant subsection retrieval for law domain question answer system

Abstract

Intelligent and instinctive legal document subsection information retrieval system is much needed for the appropriate jurisprudential system. To satisfy the law stakeholders’ need, the system should be able to deal with the semantics of law domain content. In this chapter, a sophisticated legal Question-Answer (QA) system is developed specifically for law domain which will be able to retrieve the relevant and best suitable document for any specific law domain queries posted by users’. Legal QA system is developed with the help of two computational areas— Natural Language Processing and Information Retrieval. This system is developed in an amenable way to retrieve the relevant subsection in accordance with the legal terminology embedded inquiry entered by the user. Syntactic and semantic analysis of legal documents followed by query processing helps in retrieving inferences from the knowledge base to answer the query. In our research, various models have been analyzed in the opinion of the document matching threshold value. Satisfactory results are obtained by 0.5 threshold value.