Elephant: Sequence Labeling for Word and Sentence Segmentation
Résumé
Tokenization is widely regarded as a solved problem due to the high accuracy that rule-based tokenizers achieve. But rule-based tokenizers are hard to maintain and their rules language specific. We show that high-accuracy word and sentence segmentation can be achieved by using supervised sequence labeling on the character level combined with unsupervised feature learning. We evaluated our method on three languages and obtained error rates of 0.27 ‰ (English), 0.35 ‰ (Dutch) and 0.76 ‰ (Italian) for our best models .
Origine : Fichiers produits par l'(les) auteur(s)
Loading...