Legal Text Classification on Legal Text Classification (75/25)
97.26AccuracyCNN + FastText + Lemmatisation
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| CNN + FastText + Lemmatisation2026.04 | 97.26 | 97.28 | 97.26 | 96.82 | |
| BERTModel Variant=Base, Training Mode=Fine-tuned2026.04 | 97.12 | 97.05 | 97.12 | 96.88 | |
| Legal-BERTSource=reported2026.04 | 96.9 | — | — | — | |
| BERT-CNNSource=reported2026.04 | 96.8 | — | — | — | |
| LSTMEmbedding=FastText2026.04 | 95.68 | 95.4 | 95.68 | 95.1 | |
| CNNEmbedding=Random Embedding2026.04 | 93.51 | 93.2 | 93.51 | 92.88 | |
| KNNRepresentation=TF-IDF2026.04 | 89.42 | 88.95 | 89.42 | 88.6 |