Discriminative Optimization of the Lexical Model
نویسندگان
چکیده
We publish first experiments on a new approach for training unigram prior probabilities of pronunciation variants using the Discriminative Model Combination (DMC) framework. A specific analysis of the approach is provided together with an error analysis before and after the DMC-training of the pronunciation priors.
منابع مشابه
Lexical Features for Statistical Machine Translation
Title of dissertation: LEXICAL FEATURES FOR STATISTICAL MACHINE TRANSLATION Jacob Devlin, Master of Science, 2009 Dissertation directed by: Professor Bonnie Dorr Department of Computer Science In modern phrasal and hierarchical statistical machine translation systems, two major features model translation: rule translation probabilities and lexical smoothing scores. The rule translation probabil...
متن کاملLexicon Optimization for Automatic Speech Recognition based on Discriminative Learning
In agglutinative languages such as Japanese and Uyghur, selection of lexical unit is not obvious and one of the important issues in designing language model for automatic speech recognition (ASR). In this paper, we propose a discriminative learning method to select word entries which would reduce the word error rate (WER). We define an evaluation function for each word by a set of features and ...
متن کاملJoint optimization on decoding graphs using minimum classification error criterion
Motivated by the inherent correlation between the speech features and their lexical words, we propose in this paper a new framework for learning the parameters of the corresponding acoustic and language models jointly. The proposed framework is based on discriminative training of the models’ parameters using minimum classification error criterion. To verify the effectiveness of the proposed fra...
متن کاملPotts Model on the Case Fillers for Word Sense Disambiguation
We propose a new method for word sense disambiguation for verbs. In our method, sense-dependent selectional preference of verbs is obtained through the probabilistic model on the lexical network. The meanfield approximation is employed to compute the state of the lexical network. The outcome of the computation is used as features for discriminative classifiers. The method is evaluated on the da...
متن کاملA Discriminative Lexicon Model for Complex Morphology
This paper describes successful applications of discriminative lexicon models to the statistical machine translation (SMT) systems into morphologically complex languages. We extend the previous work on discriminatively trained lexicon models to include more contextual information in making lexical selection decisions by building a single global log-linear model of translation selection. In offl...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2000