Topic-Based Legislation Network Mapping with Chatgpt and NEO4J

Legislation is an essential element of Indonesia's legal system, but the complexity and growing volume of regulations pose challenges in the accessibility and public understanding of legal information. The generally unstructured format of regulatory documents, such as Portable Document Format (PDF), makes it difficult to find specific information and analyze the linkages between regulations. This research aims to develop Legal Network Analytics (LNA), a web-based legal information system that maps linkages between legislations based on the same or similar topics. LNA utilizes the GPT-3.5-Turbo language model provided by ChatGPT to extract topics from regulatory documents and uses the Neo4j graph database to model and visualize relationships between regulations. The results show that Legal Network Analytics can identify and extract relevant topics from regulatory documents, as well as map and visualize regulatory networks interactively. Thus, Legal Network Analytics can be an effective tool for the public, legal practitioners, policymakers, and academics to understand and optimally utilize legal information.