Approaches for information retrieval in legal documents

Abstract

With crimes increasing at an alarming rate, it becomes essential to impart justice to the victims readily. To come to a final decision, lawyers need to study several previous judgments for research purposes. Reducing the time spent on research can speed up the judicial process drastically. The time consumption mostly happens in two areas — searching for the right document and understanding that document. To start with, being able to get hold of the appropriate judgments or other legal documents is the most essential task for any legal professional, especially lawyers. Once a document is obtained, the next most integral and inevitable task is to read and re-read it, and come to necessary as well as needful conclusions after a comprehensive analysis. To resolve the first issue, there is a need for an efficient search system which can provide searching options based upon multiple views. This system is an effort at improving the search for users by providing them with search options based upon either the semantics of the word or based upon the IPC sections. It is important that laymen can access all the related judgments by entering just one keyword or phrase without bothering about the legal jargon. Post retrieval of documents, the lengthy texts have to be scrutinized for meaningful inferences. To reduce the time spent in reading texts, we intend to present the information in the judgments visually through semantic networks. Lawyers will be benefitted by this system because it will enable them to skip the complexity of the often verbose language of the legal documents. This IR System provides the features of semantic and IPC section based search to users, deriving information from semantic networks that are representative of the documents, so that a more efficient search system on legal documents can be put in place.