SP-23.1

Speech Recognition Experiments Using Multi-Span Statistical Language Models
Jerome R Bellegarda (Apple Computer)

A multi-span framework was recently proposed to integrate the various constraints, both local and global, that are present in the language. In this approach, local constraints are captured via n-gram language modeling, while global constraints are taken into account through the use of latent semantic analysis. The performance of the resulting multi-span language models, as measured by perplexity, has been shown to compare favorably with the corresponding n-gram performance. This paper reports on actual speech recognition experiments, and shows that word error rate is also substantially reduced. On a subset of the Wall Street Journal speaker-independent, 20,000-word vocabulary, continuous speech task, the multi-span framework resulted in a reduction in average word error rate of up to 17%.

SP-23.2

Spoken Language Variation over Time and State in a Natural Spoken Dialog System
Allen L Gorin, Giuseppe Riccardi (AT&T-Labs Research)

We are interested in adaptive spoken dialog system for automated services. Peoples' spoken language usage varies over time for a fixed task, and furthermore varies depending on the state of the dialog. We will characterize and quantify this variation based on a database of 20K user-transactions with AT&T's experimental " How May I Help You" spoken dialog system. We then report on a language adaptation algorithm which was used to train state-dependent ASR language models, experimentally evaluating their improved performance with respect to word accuracy and perplexity.

SP-23.3

COMPARATIVE EVALUATION OF SPOKEN CORPORA ACQUIRED BY PRESENTATION OF VISUAL SCENARIOS AND TEXTUAL DESCRIPTIONS
Demetrio Aiello, Cristina Delogu (Fondazione Ugo Bordoni), Renato De Mori (Universit� d'Avignon et des Pays de Vaucluse), Andrea Di Carlo, Marina Nisi, Silvia Tummeacciu (Fondazione Ugo Bordoni)

The paper describes a system, in JAVA, for written and visual scenario generation used to collect speech corpora in the framework of a Tourism Information System. Experimental evidence shows that the corpus generated with visual scenarios has a higher perplexity and a richer vocabulary than the corpus generated using the same conceptual derivations to produce textual scenarios. Furthermore, there is evidence that textual scenarios influence speakers in the choice of the lexicon used to express the concepts more than visual scenarios.

SP-23.4

Using smoothed K-TSS language models in Continuous Speech Recognition systems
Amparo Varona, Ines Torres (Departamento de Electricidad y Electronica. Universidad del Pais Vasco.)

A syntactic approach of the well-known N-grams models, the K-Testable Language in the Strict Sense (K-TSS), is used in this work to be integrated in a Continuous Speech Recognition (CSR) system. The use of smoothed K-TSS regular grammars allowed to obtain a deterministic Stochastic Finite State Automaton (SFSA) integrating K k-TSS models into a self-contained model. An efficient representation of the whole model in a simple array of and adequate size is proposed. This structure can be easily handled at decoding time by a simple search function through the array. This formulation strongly reduced the number of parameters to be managed and thus the computing complexity of the model. An experimental evaluation of the proposed SFSA representation was carried out over an Spanish recognition task. These experiments showed important memory saving to allocate K-TSS Language models, more important for higher values of K. They also showed that the decoding time did not meaningfully increased when K did. The lower word error rates for the Spanish task tested were achieved for K=4 and 5. As a consequence the ability of this syntactic approach of the N-grams to be well integrated in a CSR system, even for high values of K, has been established.

SP-23.5

Interfacing a CDG Parser with an HMM Word Recognizer Using Word Graphs
Mary P Harper, Michael T Johnson, Leah H Jamieson, Stephen A Hockema, Christopher M White (Purdue University, School of Electrical and Computer Engineering)

In this paper, we describe a prototype spoken language system that loosely integrates a speech recognition component based on hidden Markov models with a constraint dependency grammar (CDG) parser using a word graph to pass sentence candidates between the two modules. This loosely coupled system was able to improve the sentence selection accuracy and concept accuracy over the level achieved by the acoustic module with a stochastic grammar. Timing profiles suggest that a tighter coupling of the modules could reduce parsing times of the system, as could the development of better acoustic models and tighter parsing constraints for conjunctions.

SP-23.6

An automatic acquisition method of statistic finite-state automaton for sentences
Motoyuki SUZUKI, Shozo MAKINO (Computer Center / Graduate School of Information Sciences, TOHOKU University), Hirotomo ASO (Graduate School of Enginieering, TOHOKU University)

Statistic language models obtained from a large number of training samples play an important role in speech recognition. In order to obtain higher recognition performance, we should introduce long distance correlations between words. However, traditional statistic language models such as word n-grams and ergodic HMMs are insufficient for expressing long distance correlations between words. In this paper, we propose an acquisition method for a language model based on HMnet taking into consideration long distance correlations and word location.

SP-23.7

Robust Dialogue-State Dependent Language Modeling Using Leaving-One-Out
Frank Wessel, Andrea Baader (RWTH Aachen)

The use of dialogue-state dependent language models in automatic inquiry systems can improve speech recognition and understanding if a reasonable prediction of the dialogue-state is feasible. In this paper, the dialogue-state is defined as the set of parameters which are contained in the system prompt. For each dialogue-state a separate language model is constructed. In order to obtain robust language models despite the small amount of training data we propose to interpolate all of the dialogue-state dependent language models linearly for each dialogue-state and to train the large number of resulting interpolation weights with the EM-Algorithm in combination with Leaving-One-Out. We present experimental results on a small Dutch corpus which has been recorded in the Netherlands with a train timetable information system and show that the perplexity and the word error rate can be reduced significantly.

SP-23.8

On-line Algorithms for Combining Language Models
Adam Kalai, Stanley Chen, Avrim Blum, Ronald Rosenfeld (Carnegie Mellon University)

Multiple language models are combined for many tasks in language modeling, such as domain and topic adaptation. In this work, we compare on-line algorithms from machine learning to existing algorithms for combining language models. On-line algorithms developed for this problem have parameters that are updated dynamically to adapt to a data set during evaluation. On-line analysis provides guarantees that these algorithms will perform nearly as well as the best model chosen in hindsight from a large class of models, e.g., the set of all static mixtures. We describe several on-line algorithms and present results comparing these techniques with existing language modeling combination methods on the task of domain adaptation. We demonstrate that, in some situations, on-line techniques can significantly outperform static mixtures (by over 10% in terms of perplexity) and are especially effective when the nature of the test data is unknown or changes over time.

< SP-22 SP-24 >

Last Update: February 4, 1999 Ingo Höntsch