| India's judiciary faces a backlog exceeding 44 million pending cases, with a significant fraction attributable to petitions filed in inappropriate judicial forums due to the absence of automated routing guidance at the point of electronic filing. We present LeX-Route, an automated petition routing system that classifies legal petitions into six judicial court categories using a hybrid feature representation combining semantic embeddings from LegalBERT, structured legal code multi-hot vectors (IPC sections and statutory Acts), and eight argument-mined scalar features. Evaluated on 2,467 Supreme Court petitions from the Indian Legal Documents Corpus (ILDC) under stratified 5-fold cross-validation with SMOTE-based oversampling, LightGBM achieves 92.50\% accuracy and 92.48\% macro-F1. End-to-end fine-tuned transformers (InLegalBERT: 66.88\%, LegalBERT fine-tuned: 65.99\%) substantially underperform despite dedicated Indian legal pre-training, confirming that petition routing is fundamentally a structured-signal problem where explicit statutory references dominate over contextual embeddings. While a tuned MLP reaches 93.32\% accuracy, LightGBM is the recommended deployment model owing to its interpretable feature importance rankings, satisfying the explainability requirements critical for judicial decision-support systems. To our knowledge, this is the first machine learning study targeting prospective petition routing in India's judiciary, offering a practical path toward reducing transfer delays through eCourts integration. |
*** Title, author list and abstract as submitted during Camera-Ready version delivery. Small changes that may have occurred during processing by Springer may not appear in this window.