Sinhala legal text summarizing system

Abstract

Legal domain have large amount of important, confidential information in different categories. Legal documents/ legal Acts contain large volumes, different vocabulary and text with multiple different meanings. This research is about Sinhala Public Law Acts which has complex vocabulary, citations and complex content. When readers go through the legal documents, they face problems like understanding the content because there are many paragraphs describing minor details in various aspects. This could be a very tiring task. Therefore they will not be able to concentrate, may skip content and consequently will miss the important parts. Additionally because of its complex content and many pages, the reader will have a lazy mind to read the whole document. By not reading the whole content and not understanding it properly, it may push the reader to the wrong conclusions. Therefore having summary of legal document with understandable vocabulary will help reader to overcome such difficulties This paper describes a system for legal professional and laymen to get summaries of Sinhala legal documents using deep learning techniques in natural language processing based on abstractive text summarizing method.