Using generic predictive models for slot values in language modeling

作者: David M. Chickering , Timothy S. Paek

DOI:

关键词: ArgumentLanguage modelComputer scienceData storeDummy variableGeneric propertyDistribution (mathematics)Data miningSet (abstract data type)Value (computer science)

摘要: A generic predictive argument model that can be applied to a set of shot values predict target slot value is provided. The whether or not particular item the intended user command given various features. prediction for each then normalized infer distribution over all items. For any (e.g., contacts), number binary variable s are created indicate specific was target. value, input features employed corresponding variable. These properties contact “instantiated” based o n contact-specific features). stored in data store.

参考文章(70)
Eric Horvitz, Tim Paek, Optimizing Automated Call Routing by Integrating Spoken Dialog Models with Queuing Models north american chapter of the association for computational linguistics. pp. 41- 48 ,(2004)
Katunobu Itou, Atsushi Fujii, Tetsuya Ishikawa, Tomoyosi Akiba, Selective back-off smoothing for incorporating grammatical constraints into the n-gram language model. conference of the international speech communication association. ,(2002)
Nobuyuki Washio, Grammar update system and method ,(2003)
Johan Schalkwyk, Michael S. Phillips, Speech recognition with dynamic grammars ,(2001)