Answering legal questions by mining reference information

Abstract

This paper presents a study on exploiting reference information to build a question answering system restricted to the legal domain. Most previous research focuses on answering legal questions whose answers can be found in one document (The term ‘documents’ corresponds to articles, paragraphs, items, or sub-items according to the naming rules used in the legal domain.) without using reference information. However, there are many legal questions whose answers could not be found without linking information from multiple documents. This connection is represented by explicit or implicit references. To the best of our knowledge, this type of questions is not adequately considered in previous work. To cope with them, we propose a novel approach which allow us to exploit the reference information among legal documents to find answers. This approach also uses requisite-effectuation structures of legal sentences and some effective similarity measures to support finding correct answers without training data. The experimental results showed that the proposed method is quite effective and outperform a traditional QA method, which does not use reference information.