Automatic Construction of a Legal Citation Graph from 100 Million Ukrainian Court Decisions
Half a billion citation edges extracted from 100.7 million Ukrainian court decisions reveal that judicial citation structure encodes legal domain boundaries without supervision and predicts future legislative importance with near-perfect accuracy (AUC = 0.9984).
Automatic Construction of a Legal Citation Graph from 100 Million Ukrainian Court Decisions
Large-Scale Extraction, Topological Analysis, and Ontology-Driven Clustering
Volodymyr Ovcharov — LEX AI LLC, Kyiv, Ukraine
Abstract
Half a billion citation edges extracted from 100.7 million Ukrainian court decisions reveal that judicial citation structure encodes legal domain boundaries without supervision and predicts future legislative importance with near-perfect accuracy (AUC = 0.9984). 502 million citation links extracted via regex with precision 1.00. Power-law α=1.57, cross-domain bridging 73.1%, NMI 0.83–0.86 stable ontology, 2022 war entropy spike.