Structured knowledge graphs improve legal document search accuracy
Verifiable Retrieval for Legal Reasoning through Structure-driven Knowledge Graph
Computation and Language
Summary
Finding precise legal information is hard because laws are complex and connected. The authors show that building a map of legal rules using a knowledge graph helps computer programs find important legal citations more accurately. This approach makes it easier to check where answers come from, helping with transparency and trust. It works better than just searching chunks of documents without considering their structure.
What this means in practice
- •For legal technology developers: Build tools that retrieve and verify citations from complex legal codes using structured knowledge graphs.$Commercial implications: Software developers can create commercial legal research platforms offering more reliable, transparent citation retrieval based on this structured approach.
- •For compliance teams: Use accurate citation retrieval to verify legal references in compliance documents automatically, reducing errors.
Authors
Ibrahim Siddig, Sviatoslav Tugeev, Munir Georges
Abstract
Any automation within reliability-critical contexts such as medicine or law has to be transparent and verifiable. Despite their advantages, current LLM-based systems do not meet these requirements due to their inherent black-box nature. Retrieval Augmented Generation (RAG) attempts to ground the generative process in the relevant interpretable documents, thereby reducing the risk of AI 'hallucinations' and potentially enabling verification of the initial generation context. However, as the number and size of potential sources increases, naive retrieval approaches tend to perform worse, ignoring critical context elements. In this paper we aim to tackle this problem in the context of German Legal Code. Using structure-driven parsing to create a Knowledge Graph (KG) for the source document navigation and retrieval, we compare several KG-informed chunking and retrieval strategies and evaluate their effect on the accuracy of retrieved citations.