Temporal Concept Drift in Legal Judgment Prediction: Neural Baselines Across Three Epochs of Ukrainian Court Decisions

In three linesStudy of temporal concept drift in legal NLP on 428K Ukrainian court decisions (2008-2026). Four transformer models (XLM-RoBERTa, legal variants) show severe forward degradation (−27.2 pp macro-F1) but robust backward transfer. Chronological continual learning eliminates catastrophic forgetting.

Summary generated by Claude — human-verified