Out-of-vocabulary (OOV) words create serious problems for automatic speech recognition (ASR) systems. Not only are they miss-recognized as in-vocabulary (IV) words with similar phonetics, but the error also causes further errors in nearby words. Language models (LMs) for most open vocabulary ASR systems treat OOV words as a single entity, ignoring the linguistic information. In this paper we present a class-based n-gram LM that is able to deal with OOV words by treating each of them individually without retraining all the LM parameters. OOV words are assigned to IV classes consisting of similar semantic meanings for IV words. The World Wide Web is used to acquire additional data for finding the relation between the OOV and IV words. An evaluation based on adjusted perplexity and word-error-rate was carried out on the Wall Street Journal corpus. The result suggests the preference of the use of multiple classes for OOV words, instead of one unknown class.

This paper incorporates statistical language models into an on-line handwriting recognition system for devices with limited memory and computational resources. The objective is to minimize the error recognition rate by taking into account the sentence context to disambiguate poorly written texts. Probabilistic word n-grams have been first investigated, then to fight the curse of dimensionality problem induced by such an approach and to decrease significantly the size of the language model an extension to class-based n-grams has been achieved. In the latter case, the classes result either from a syntactic criterion or a contextual criteria. Finally, a composite model is proposed; it combines both previous kinds of classes and exhibits superior performances compared with the word n-grams model. We report on many experiments involving different European languages (English, French, and Italian), they are related either to language model evaluation based on the classical perplexity measurement on test text corpora but also on the evolution of the word error rate on test handwritten databases. These experiments show that the proposed approach significantly improves on state-of-the-art n-gram models, and that its integration into an on-line handwriting recognition system demonstrates a substantial performance improvement.

In this paper, we discuss the construction of a large in-car spoken dialogue corpus and the result of its analysis. We have developed a system specially built into a Data Collection Vehicle (DCV) which supports the synchronous recording of multichannel audio data from 16 microphones that can be placed in flexible positions, multichannel video data from 3 cameras, and vehicle related data. Multimedia data has been collected for three sessions of spoken dialogue with different modes of navigation, during approximately a 60 minute drive by each of 800 subjects. We have characterized the collected dialogues across the three sessions. Some characteristics such as sentence complexity and SNR are found to differ significantly among the sessions. Linear regression analysis results also clarify the relative importance of various corpus characteristics.

For the purpose of automatic speech recognition, language models (LMs) are used to predict possible succeeding words for a given partial word sequence and thereby to reduce the search space. In this paper several kinds of stochastic language models (SLMs) are evaluated-bigram, trigram, hidden Markov model (HMM), bigram-HMM, stochastic context-free grammar (SCFG) and hand-written Bunsetsu Grammar. To compare the predictive power of these SLMs, the evaluation was conducted from two points of views: (1) relationship between the number of model parameters and entropy, (2) predictive rate of succeeding part of speech (POS) and succeeding word. We propose a new type of bigram-HMM and compare it with the other models. Two kinds of approximations are tried and examined through experiments. Results based on both of English Brown-Corpus and Japanese ATR dialog database showed that the extended bigram-HMM had better performance than the others and was more suitable to be a language model.