Information Retrieval Based Legal Search System

Abstract

Calculating the similarity between two legal documents to find similar legal judgments is an important challenge in legal information. Efficiently computing this similarity by expanding widely used information retrieval and search engine techniques has practical applications in a number of tasks, like locating pertinent prior cases for a specific case document. Programmed data recovery frameworks or reports are the main parts of today’s selected emotional support networks or web indexes to reduce data overload. Investigating methodologies to work on the presentation of report recovery frameworks and web search tools is a working area of research. Various methods have been pro- posed in this research paper to explore ways to search the common law system for cases with a similar outcome. Building a legal decision support system is intended to increase efficiency by assisting stakeholders—including judges and attorneys—in finding related rulings promptly. In order to prepare arguments, a lawyer typically has to review earlier decisions that are comparable to (or pertinent to) the current case. The attorney examines the judgement database to discover similar judgements. Legal rulings are complex in nature and refer to other judgments. For this, proper techniques are needed for quality analysis of judgments and correct deductions from them. A proper analysis of several types of similarity measures, such as all-term-based similarity methods, legal terms, co-citations, and bibliographic links, performed to look for comparable conclusions. According to experimental findings, the law term similarity approach outperforms all term cosine similarity methods. The out- comes also demonstrate that the co-citation approach performs worse than the bibliographic linkage similarity method and improves performance over the co-citation approach. After proper analysis of various methods in this field, proper comparison can be made between documents and similar legal documents can also be easily searched based on their similarity pattern and can be used to make meaningful deductions.