Spoken Language Understanding Systems 3
HomeFull List of Titles
1: ICSLP'98 Proceedings
Keynote SpeechesText-To-Speech Synthesis 1Spoken Language Models and Dialog 1Prosody and Emotion 1Hidden Markov Model Techniques 1Speaker and Language Recognition 1Multimodal Spoken Language Processing 1Isolated Word RecognitionRobust Speech Processing in Adverse Environments 1Spoken Language Models and Dialog 2Articulatory Modelling 1Talking to Infants, Pets and LoversRobust Speech Processing in Adverse Environments 2Spoken Language Models and Dialog 3Speech Coding 1Articulatory Modelling 2Prosody and Emotion 2Neural Networks, Fuzzy and Evolutionary Methods 1Utterance Verification and Word Spotting 1 / Speaker Adaptation 1Text-To-Speech Synthesis 2Spoken Language Models and Dialog 4Human Speech Perception 1Robust Speech Processing in Adverse Environments 3Speech and Hearing Disorders 1Prosody and Emotion 3Spoken Language Understanding Systems 1Signal Processing and Speech Analysis 1Spoken Language Generation and Translation 1Spoken Language Models and Dialog 5Segmentation, Labelling and Speech Corpora 1Multimodal Spoken Language Processing 2Prosody and Emotion 4Neural Networks, Fuzzy and Evolutionary Methods 2Large Vocabulary Continuous Speech Recognition 1Speaker and Language Recognition 2Signal Processing and Speech Analysis 2Prosody and Emotion 5Robust Speech Processing in Adverse Environments 4Segmentation, Labelling and Speech Corpora 2Speech Technology Applications and Human-Machine Interface 1Large Vocabulary Continuous Speech Recognition 2Text-To-Speech Synthesis 3Language Acquisition 1Acoustic Phonetics 1Speaker Adaptation 2Speech Coding 2Hidden Markov Model Techniques 2Multilingual Perception and Recognition 1Large Vocabulary Continuous Speech Recognition 3Articulatory Modelling 3Language Acquisition 2Speaker and Language Recognition 3Text-To-Speech Synthesis 4Spoken Language Understanding Systems 4Human Speech Perception 2Large Vocabulary Continuous Speech Recognition 4Spoken Language Understanding Systems 2Signal Processing and Speech Analysis 3Human Speech Perception 3Speaker Adaptation 3Spoken Language Understanding Systems 3Multimodal Spoken Language Processing 3Acoustic Phonetics 2Large Vocabulary Continuous Speech Recognition 5Speech Coding 3Language Acquisition 3 / Multilingual Perception and Recognition 2Segmentation, Labelling and Speech Corpora 3Text-To-Speech Synthesis 5Spoken Language Generation and Translation 2Human Speech Perception 4Robust Speech Processing in Adverse Environments 5Text-To-Speech Synthesis 6Speech Technology Applications and Human-Machine Interface 2Prosody and Emotion 6Hidden Markov Model Techniques 3Speech and Hearing Disorders 2 / Speech Processing for the Speech and Hearing Impaired 1Human Speech ProductionSegmentation, Labelling and Speech Corpora 4Speaker and Language Recognition 4Speech Technology Applications and Human-Machine Interface 3Utterance Verification and Word Spotting 2Large Vocabulary Continuous Speech Recognition 6Neural Networks, Fuzzy and Evolutionary Methods 3Speech Processing for the Speech-Impaired and Hearing-Impaired 2Prosody and Emotion 72: SST Student Day
SST Student Day - Poster Session 1SST Student Day - Poster Session 2Author Index
A B C D E F G H I
J K L M N O P Q R
S T U V W X Y Z
Multimedia Files
Language Modeling for Content Extraction in Human-Computer Dialogues
Authors:
Wolfgang Reichl, Bell-Labs, Lucent Technologies (USA)
Bob Carpenter, Bell-Labs, Lucent Technologies (USA)
Jennifer Chu-Carroll, Bell-Labs, Lucent Technologies (USA)
Wu Chou, Bell-Labs, Lucent Technologies (USA)
Page (NA) Paper number 588
Abstract:
In this paper we discuss the role of language modeling in a novel natural language dialogue system designed to automatically route incoming customer calls. We arrive at two significant conclusions: First, standard word error rate measures do not reflect application specific requirements; highly reliable content extraction is possible with relatively high word error rates. Secondly blending human-human data with human-machine data did not improve the performance in language modeling.
SL980588.PDF (From Author)
SL980588.PDF (Rasterized)
TOP
A Language Model Combining Trigrams and Stochastic Context-Free Grammars
Authors:
John Gillett, Carnegie Mellon University (USA)
Wayne Ward, Carnegie Mellon University (USA)
Page (NA) Paper number 872
Abstract:
We propose a class trigram language model in which each class is specified by a probabilistic context-free grammar. We show how to estimate the parameters of the model, and how to smooth these estimates. We present experimental perplexity and speech recognition results.
SL980872.PDF (From Author)
SL980872.PDF (Rasterized)
TOP
Online Adaptation of Language Models in Spoken Dialogue Systems
Authors:
Bernd Souvignier, Philips Research Laboratories (Germany)
Andreas Kellner, Philips Research Laboratories (Germany)
Page (NA) Paper number 961
Abstract:
The robust estimation of language models for new applications of spoken dialogue systems often suffers from a lack of available training material. An alternative to training is to adapt initial language models to a new task by exploiting material from recognition. We investigate different methods for online-adaptation of language models. Apart from supervised and unsupervised adaptation, we look at two refined approaches: the first allows multiple hypotheses from N-best lists for adaptation and the second uses confidence measures to reject unreliably recognized sentences. We apply adaptation both to the language model used in the recognizer to focus the beam search and to the stochastic language understanding grammar. It turns out that the understanding grammar can be improved quite significantly using N-best lists or confidence measures, whereas unsupervised adaptation may even result in a deterioration of the system. The language model used in the recognizer is also improved very satisfactory.
SL980961.PDF (From Author)
SL980961.PDF (Rasterized)
TOP
Language Model Adaptation for Spoken Language Systems
Authors:
Giuseppe Riccardi, AT&T-Labs Research (USA)
Alexandros Potamianos, AT&T-Labs Research (USA)
Shrikanth Narayanan, AT&T-Labs Research (USA)
Page (NA) Paper number 1052
Abstract:
In a human-machine interaction (dialog) the statistical language variations are large among different stages of the dialog and across different speakers. Moreover, spoken dialog systems require extensive training data for training stochastic language models. In this paper we address the problem of open-vocabulary language models allowing the user for any possible response at each stage of the dialog. We propose a novel off-line adaptation of stochastic language models effective for their generalization (open-vocabulary) and selective (dialog context) properties. We outline the integration of the finite state dialog and the language model adaptation algorithm. The performance of the speech recognition and understanding language models are evaluated with the Carmen Sandiego multimodal computer game. The new language models give an overall understanding error rate reduction of 44% over the baseline system.
SL981052.PDF (From Author)
SL981052.PDF (Rasterized)
TOP
Detecting Topic Shifts Using a Cache Memory
Authors:
Brigitte Bigi, LIA CERI-IUP University of Avignon (France)
Renato De Mori, LIA CERI-IUP University of Avignon (France)
Marc El-Beze, LIA CERI-IUP University of Avignon (France)
Thierry Spriet, LIA CERI-IUP University of Avignon (France)
Page (NA) Paper number 77
Abstract:
The use of cache memories and symmetric Kullback-Leibler distances is proposed for topic classification and topic-shift detection. Experiments with a large corpus of articles from the French newspaper "Le Monde show tangible advantages when different models are combined with a suitable strategy. Experimental results show that different strategies for topic shift detection have to be used depending on whether high recall or high precision are sought. Furthermore, methods based on topic independent distributions provide complementary candidates with respect to the use of topic-dependent distributions leading to an increase in recall with a minor loss in precision.
SL980077.PDF (From Author)
SL980077.PDF (Rasterized)
TOP
A Discourse Coding Scheme for Conversational Spanish
Authors:
Lori Levin, Carnegie Mellon University (USA)
Ann Thymé-Gobbel, Natural Speech Technologies (USA)
Alon Lavie, Carnegie Mellon University (USA)
Klaus Ries, Carnegie Mellon University (USA)
Klaus Zechner, Carnegie Mellon University (USA)
Page (NA) Paper number 1000
Abstract:
This paper describes a 3-level manual discourse coding scheme that we have devised for manual tagging of the CallHome Spanish (CHS) and CallFriend Spanish (CFS) databases used in the CLARITY project. The goal of CLARITY is to explore the use of discourse structure in understanding conversational speech. The project combines empirical methods for dialogue processing with state-of-the art LVCSR (using the JANUS recognizer). The three levels of the coding scheme are (1) a speech act level consisting of a tag set extended from DAMSL and Switchboard; (2) dialogue game level defined by initiative and intention; and (3) an activity level defined within topic units. The manually tagged dialogues are used to train automatic classifiers. We present preliminary results for automatic speech act classification and topic boundary identification and inter-coder speech act confusion matrices.