A spoken dialogue system for German intercity train timetable inquiries

Eckert, Wieland; Kuhn, Tobias; Niemann, Heinrich; Rieck, S.; Scheuer, Alfred Q.; Schukat-Talamazzini, Ernst Günter

doi:10.21437/eurospeech.1993-428

Cited by 20 publications

(5 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Experiments were performed on the EVAR corpus of spontaneous speech data collected by our spoken dialog system [3,13] which is able to answer inquiries about German Intercity train connections. The data were divided into a training sample, a validation sample and a test sample (Table 1).…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…In other applications like information retrieval over the telephone the user might not even know that the system misrecognized because of an OOV word. So if a user asks our train timetable inquiry system [3] "I want to go from Sussex" and Sussex is not in the lexicon, the system might recognize I want to go at six and it might respond with "You want to leave at six o'clock. Where do you want to go?".…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A category based approach for recognition of out-of-vocabulary words

Gallwitz¹,

Nöth²,

Niemann³

1996

4th International Conference on Spoken Language Processing (ICSLP 1996)

Self Cite

View full text Add to dashboard Cite

In almost all applications of automatic speech recognition, especially in spontaneous speech tasks, the recognizer vocabulary cannot cover all occurring words. There is always a significant amount of out-of-vocabulary words even when the vocabulary size is very large. In this paper we present a new approach for the integration of out-of-vocabulary words into statistical language models. We use category information for all words in the training corpus to define a function that gives an approximation of the out-of-vocabulary word emission probability for each word category. This information is integrated into the language models. Although we use a simple acoustic model for out-of-vocabulary words, we achieve a 6% reduction of word error rate on spontaneous speech data with about 5% out-of-vocabulary rate.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A category based approach for recognition of out-of-vocabulary words

Gallwitz¹,

Nöth²,

Niemann³

1996

4th International Conference on Spoken Language Processing (ICSLP 1996)

Self Cite

View full text Add to dashboard Cite

show abstract

“…While WA e v aluates the performance of the speech recognition component, the language understanding capabilities of a system can be judged by concept accuracy (CA). 3 This approach is based on the assumption that the main task of the linguistic processor in a spoken dialogue system is to extract the propositional content from the spoken utterance. Furthermore, it is assumed that this propositional content can be represented as a list of semantic units (SU) taking the form of attribute-value pairs.…”

Section: Concept Accuracymentioning

confidence: 99%

“…Given such semantic reference answers in form of task parameter-value pairs the performance of a speech understanding component can be measured in analogy with the 3 In [10] a similar measure was called information content.…”

Section: Concept Accuracymentioning

confidence: 99%

“…[6], [11], [2]). This paper describes our approach to automatic evaluation of both the recognition and the understanding capabilities of a s p o k en dialogue system for train time table inquiries [3]. Such a n i n tegrated evaluation environment allows a systematic investigation of the relationship between recognition and understanding performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Towards understanding spontaneous speech: word accuracy vs. concept accuracy

Boros¹,

Eckert²,

Gallwitz³

et al. 1996

4th International Conference on Spoken Language Processing (ICSLP 1996)

Self Cite

View full text Add to dashboard Cite

In this paper we describe an approach to automatic evaluation of both the speech recognition and understanding capabilities of a spoken dialogue system for train time table information. We use word a c curacy for recognition and concept accuracy for understanding performance judgement. Both measures are calculated by comparing these modules' output with a correct reference answer. We report evaluation results for a spontaneous speech corpus with about 10000 utterances. We observed a nearly linear relationship between word accuracy and concept accuracy.

show abstract

Analysis of Different Dialog Strategies in the Slovenian Spoken Dialog System

Ipšić

Pavešić

1999

Text, Speech and Dialogue

View full text Add to dashboard Cite

A spoken dialogue system for German intercity train timetable inquiries

Cited by 20 publications

References 0 publications

A category based approach for recognition of out-of-vocabulary words

A category based approach for recognition of out-of-vocabulary words

Towards understanding spontaneous speech: word accuracy vs. concept accuracy

Analysis of Different Dialog Strategies in the Slovenian Spoken Dialog System

Contact Info

Product

Resources

About