Hybrid Topic-Semantic Labeling and Graph Embeddings for Unsupervised Legal Document Clustering

arXiv:2509.00990 · stat.ML, cs.CL, cs.LG · Submitted 2025-08-31 · Read on arXiv

stat.ML, cs.CL, cs.LG

Submitted: 2025-08-31

Updated: 2026-10-17

Comments: 21 pages, 11 figures, 5 tables. Major revision (v2): formalizes Topic-Graph Spherical Fusion(TGSF) with a connected multi-relational document-topic graph and L2 spherical fusion; expands evaluation across LEDGAR, ACORD, and CUAD with unimodal ablations; adds 648 factorial grid runs and 720 Bayesian TPE and NSGA-II Pareto optimization trials with fANOVA analysis

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Embeddings — 7× in this paper · explained in 2 episodes
• Clustering — 2× in this paper · explained in 2 episodes
• Adaptation — 1× in this paper · explained in 1 episode
• Human-in-the-Loop — 1× in this paper · explained in 3 episodes
• Hybrid Approach — 1× in this paper · explained in 4 episodes
• Hyperparameters — 1× in this paper · explained in 2 episodes
• Topic Modeling — 1× in this paper · explained in 2 episodes

Related papers

• Behavior of prediction performance metrics with rare events
• Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
• A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
• Efficient quadratic entropy with distance sketches
• Embedding-Bias in Conditional Independence Testing
• From Log-Odds to Shapley Values: An Explanatory Geometry for the Weighted Naive Bayes Classifier