Improving back-off models with bag of words and hollow-grams - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2010

Improving back-off models with bag of words and hollow-grams

Raphaël Rubino
  • Fonction : Auteur
  • PersonId : 776116
  • IdRef : 172390621
Georges Linarès

Résumé

Classical n-grams models lack robustness on unseen events. The literature suggests several smoothing methods: empirically , the most effective of these is the modified Kneser-Ney approach. We propose to improve this back-off model: our method boils down to back-off value reordering, according to the mutual information of the words, and to a new hollow-gram model. Results show that our back-off model yields significant improvements to the baseline, based on the modified Kneser-Ney back-off. We obtain a 0.6% absolute word error rate improvement without acoustic adaptation, and 0.4% after adaptation with a 3xRT ASR system.
Fichier principal
Vignette du fichier
Improving_back-off_models_with_bag_of_words_and_ho.pdf (260.87 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01318103 , version 1 (29-10-2016)

Identifiants

  • HAL Id : hal-01318103 , version 1

Citer

Benjamin Lecouteux, Raphaël Rubino, Georges Linarès. Improving back-off models with bag of words and hollow-grams. INTERSPEECH, Sep 2010, Makuhari, Japan. ⟨hal-01318103⟩

Collections

UNIV-AVIGNON LIA
46 Consultations
31 Téléchargements

Partager

Gmail Facebook X LinkedIn More