Spoken language understanding

Mori, Renato De; Béchet, Frédéric; Hakkani‐Tür, Dilek; McTear, Michael F.; Riccardi, Giuseppe; Tür, Gökhan

doi:10.1109/msp.2008.918413

Cited by 105 publications

(49 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Examples of statistical approaches to various components of a dialogue system can be found in (Levin and Pieraccini, 1997;Jurafsky and Martin, 2008;De Mori et al, 2008;Thomson and Young, 2010;Lemon and Pietquin, 2012; Figure 1: Architecture of a modular Spoken Dialoug System. et al, 2013;.…”

Section: Introductionmentioning

confidence: 99%

PyDial: A Multi-domain Statistical Dialogue System Toolkit

Ultes¹,

Rojas-Barahona²,

Su³

et al. 2017

Proceedings of ACL 2017, System Demonstrations

130

View full text Add to dashboard Cite

Statistical Spoken Dialogue Systems have been around for many years. However, access to these systems has always been difficult as there is still no publicly available end-to-end system implementation. To alleviate this, we present PyDial, an opensource end-to-end statistical spoken dialogue system toolkit which provides implementations of statistical approaches for all dialogue system modules. Moreover, it has been extended to provide multidomain conversational functionality. It offers easy configuration, easy extensibility, and domain-independent implementations of the respective dialogue system modules. The toolkit is available for download under the Apache 2.0 license.

show abstract

Section: Introductionmentioning

confidence: 99%

PyDial: A Multi-domain Statistical Dialogue System Toolkit

Ultes¹,

Rojas-Barahona²,

Su³

et al. 2017

Proceedings of ACL 2017, System Demonstrations

130

View full text Add to dashboard Cite

show abstract

“…The task-oriented SDS is usually domain-specific. The system creators first map the user utterances into semantic frames that contain domain-specific slots and intents using spoken language understanding (SLU) (De Mori et al, 2008). Then a set of domain-specific dialog state variables is tracked to retain the context information over turns (Williams et al, 2013).…”

Section: Introductionmentioning

confidence: 99%

Generative Encoder-Decoder Models for Task-Oriented Spoken Dialog Systems with Chatting Capability

Zhao¹,

Lu²,

Lee³

et al. 2017

Proceedings of the 18th Annual SIGdial Meeting on Discourse and Dialogue

View full text Add to dashboard Cite

Generative encoder-decoder models offer great promise in developing domaingeneral dialog systems. However, they have mainly been applied to open-domain conversations.This paper presents a practical and novel framework for building task-oriented dialog systems based on encoder-decoder models. This framework enables encoder-decoder models to accomplish slot-value independent decision-making and interact with external databases. Moreover, this paper shows the flexibility of the proposed method by interleaving chatting capability with a slotfilling system for better out-of-domain recovery. The models were trained on both real-user data from a bus information system and human-human chat data. Results show that the proposed framework achieves good performance in both offline evaluation metrics and in task success rate with human users.

show abstract

“…The purpose of an SLU system is to interpret the meaning of a speech signal (De Mori et al, 2008). SLU systems use Automatic Speech Recognition (ASR) to convert speech signal to the text of what was spoken (hypothesis), followed by semantic meaning extraction from the ASR hypothesis using Natural Language Processing (NLP).…”

Section: Introductionmentioning

confidence: 99%

Practical Application of Domain Dependent Confidence Measurement for Spoken Language Understanding Systems

Mehrabani

Thomson²,

Stern³

2018

Proceedings of the 2018 Conference of the North American Chapter Of the Association for Computational Linguistics: Hu

View full text Add to dashboard Cite

Spoken Language Understanding (SLU), which extracts semantic information from speech, is not flawless, specially in practical applications. The reliability of the output of an SLU system can be evaluated using a semantic confidence measure. Confidence measures are a solution to improve the quality of spoken dialogue systems, by rejecting low-confidence SLU results. In this study we discuss realworld applications of confidence scoring in a customer service scenario. We build confidence models for three major types of dialogue states that are considered as different domains: how may I help you, number capture, and confirmation. Practical challenges to train domain-dependent confidence models, including data limitations, are discussed, and it is shown that feature engineering plays an important role to improve performance. We explore a wide variety of predictor features based on speech recognition, intent classification, and high-level domain knowledge, and find the combined feature set with the best rejection performance for each application.

show abstract

Spoken language understanding

Cited by 105 publications

References 22 publications

PyDial: A Multi-domain Statistical Dialogue System Toolkit

PyDial: A Multi-domain Statistical Dialogue System Toolkit

Generative Encoder-Decoder Models for Task-Oriented Spoken Dialog Systems with Chatting Capability

Practical Application of Domain Dependent Confidence Measurement for Spoken Language Understanding Systems

Contact Info

Product

Resources

About