This article proposes a new confidence measure estimated for concept hypotheses provided by a semantic language model used in the context of a dialog application. This confidence measure is based upon the ontology and more precisely, upon the semantic relations between concepts. It aims at measuring how high a concept hypothesis is related to the other hypotheses of an utterance. The semantic relation confidence measure is evaluated alone, and in combination with a classical acoustic confidence measure. The two measures are also used as parameters of a decision tree. It is shown that the two confidence measures are complementary and yield good performance in terms of cross entropy relative reduction.
Cite as: Kobus, C., Damnati, G., Delphin-Poulat, L., Mori, R.D. (2006) Exploiting semantic relations for a spoken language understanding application. Proc. Interspeech 2006, paper 1269-Tue2A2O.3, doi: 10.21437/Interspeech.2006-332
@inproceedings{kobus06_interspeech,
author={Catherine Kobus and Geraldine Damnati and Lionel Delphin-Poulat and Renato De Mori},
title={{Exploiting semantic relations for a spoken language understanding application}},
year=2006,
booktitle={Proc. Interspeech 2006},
pages={paper 1269-Tue2A2O.3},
doi={10.21437/Interspeech.2006-332}
}