Bridging Law and AI: Automating IRAC Legal Reasoning using Knowledge Base and Semi-Structured Data

This research makes pioneering contributions to the intersection of artificial intelligence and legal by systematically combining IRAC-based benchmark construction, semi-structured knowledge graph representation, and reasoning-enhanced retrieval systems for automated legal analysis. Ultimately, the integrated approach demonstrates that semi-structured legal knowledge representation can achieve significant improvements in legal reasoning quality while maintaining the interpretability required by legal professionals. The practical impact extends beyond academic contributions through our fully integrated demo system, which provides the first real-world demonstration of automated legal reasoning that preserves formal rigour while being accessible to non-experts. Ultimately, our system contributes to the broader goal of making legal reasoning more accessible, interpretable, and reliable, while establishing a replicable methodology for extending automated legal reasoning to other legal domains and jurisdictions.