Towards Intelligent Legal Consultancy based on Integrating of Ontology and Knowledge Graph and Compare with LLMs

Abstract

Nowadays the necessity for employing an AI system to support the search for legal documents is increasingly evident. In this paper, a method for integrating of ontology Legal-Onto and knowledge graph, called Legal Onto-Graph model, is studied. This integration, built upon laws and legal documents, facilitates various applications, such as question answering, document similarity, and search functionality. The proposed method is useful for representing knowledge spanning multiple documents within the legal domain. Organized as a knowledge base, the integrated model powers a Questions-Answering system tailored for legal documents. The experimental results conducted on Vietnamese Land Law demonstrate the system’s effectiveness in handling common queries, outperforming other large language models in this domain based on the criteria of an intelligent legal consultancy system.