Dialogue Strategies to Overcome Speech Recognition Errors in Form-Filling Dialogue
In a spoken dialogue system, the speech recognition performance accounts for the largest part of the overall system performance. Yet spontaneous speech recognition has an unstable performance. The proposed postprocessing method solves this problem. The state of a legacy DB can be used as an important factor for recognizing a user’s intention because form-filling dialogues tend to depend on the legacy DB. Our system uses the legacy DB and ASR result to infer the user’s intention, and the validity of the current user’s intention is verified using the inferred user’s intention. With a plan-based dialogue model, the proposed system corrected 27% of the incomplete tasks, and achieved an 89% overall task completion rate.
Keywordsdialogue strategy speech recognition word error rate form-filling dialogue sub-dialogue generation
Unable to display preview. Download preview PDF.
- 3.Cavazza, M.: An Empirical Study of Speech Recognition Errors in a Task-oriented Dialogue System. In: 2th SIGdial Workshop on Discourse and Dialogue, pp. 98–105. Association for Computational Linguistics, Denmark (2001)Google Scholar
- 4.Gorrell, G.: Language Modelling and Error Handling in Spoken Dialogue System. Licentiate thesis, Linköping University, Sweden (2003)Google Scholar
- 5.Goddeau, D., Meng, H., Polifroni, J., Seneff, S., Busayapongchai, S.: A Form-Based Dialog Manager for Spoken Language Applications. In: 4th International Conference on Spoken Language, pp. 701–705. IEEE Press, USA (1996)Google Scholar
- 7.Chu-Carroll, J., Carberry, S.: Generating information-sharing sub- dialogues in expert-user consultation. In: 14th International Conference on Artificial Intelligence, pp. 1234–1250. AAAI, UK (1995)Google Scholar
- 8.Oh, J.: The Design of Plan-based Dialogue System In Task Execution Domain. M.S. thesis, Sogang University, Korea (1999) Google Scholar
- 9.Walker, M., Passonneau, R., Boland, J.: Quantitative and Qualitative Evaluation of Darpa Communicator Spoken Dialogue Systems. In: 39th Annual Meeting of the Association for Computational Linguistics, pp. 512–522. Association for Computational Linguistics, France (2001)Google Scholar
- 10.Ahn, D., Chung, M.: One-pass Semi-dynamic Network Decoding Using a Subnetwork Caching Model for Large Vocabulary Continuous Speech Recognition. IEICE Transaction. Information and Systems E87-D(5) 5, 1164–1174 (2004)Google Scholar