On the use of linguistic consistency in systems for human-computer dialogues

Estève, Yannick; Raymond, Christian; Mori, Renato De; Janiszek, David

doi:10.1109/tsa.2003.818318

Cited by 5 publications

(2 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Further benefits can be obtained by performing MAP adaptation of a general purpose LM using augmented counts inferred from a domain specific corpus. LMs built in this way have proven to be useful also for performing sentence verification based on the consensus of recognition results obtained with different LMs (Esteve et al, 2003).…”

Section: Discussionmentioning

confidence: 99%

Data augmentation and language model adaptation using singular value decomposition

Béchet

Mori

Janiszek

2004

Pattern Recognition Letters

Self Cite

View full text Add to dashboard Cite

Section: Discussionmentioning

confidence: 99%

Data augmentation and language model adaptation using singular value decomposition

Béchet

Mori

Janiszek

2004

Pattern Recognition Letters

Self Cite

View full text Add to dashboard Cite

“…• LC, a descriptor of linguistic confidence inspired by measures proposed in [23], is the ratio, for a given word string candidate, between the number of trigrams observed in the training corpus of the LM versus the total number of trigrams in the same word string. This measure is evaluated on the best word string containing and covering the whole utterance.…”

Section: B Confidence Classifiers Formentioning

confidence: 99%

Sequential Decision Strategies for Machine Interpretation of Speech

Raymond

Béchet

Camelin

et al. 2007

IEEE Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

Abstract-Recognition errors made by automatic speech recognition (ASR) systems may not prevent the development of useful dialogue applications if the interpretation strategy has an introspection capability for evaluating the reliability of the results. This paper proposes an interpretation strategy which is particularly effective when applications are developed with a training corpus of moderate size. From the lattice of word hypotheses generated by an ASR system, a short list of conceptual structures is obtained with a set of finite state machines (FSM). Interpretation or a rejection decision is then performed by a tree-based strategy. The nodes of the tree correspond to elaboration-decision units containing a redundant set of classifiers. A decision tree based and two large margin classifiers are trained with a development set to become interpretation knowledge sources. Discriminative training of the classifiers selects linguistic and confidence-based features for contributing to a cooperative assessment of the reliability of an interpretation. Such an assessment leads to the definition of a limited number of reliability states. The probability that a proposed interpretation is correct is provided by its reliability state and transmitted to the dialogue manager. Experimental results are presented for a telephone service application.Index Terms-Confidence measures, decision strategy, speech recognition, spoken dialogue systems, spoken language understanding.

show abstract