System and method for machine-mediated human-human conversation
作者： Srinivas Bangalore
DOI:
关键词:
摘要: Disclosed herein are systems, methods, and non-transitory computer-readable storage media for processing speech. A system configured to practice the method monitors user utterances generate a conversation context. Then receives current utterance independent of non-natural language input intended trigger speech processing. The compares context similarity score, if score is above threshold, incorporates into If below discards utterance. can compare based on an n-gram distribution, perplexity threshold. Alternately, use task model
lens.org 参考文章(29)
Michael Kuperstein, Speech recognition dialog management ,(2005)
Edward Mark Tecot, Robert Chambers, Karen M. Cross, Michael Bodell, Nick Gedge, Michael Kim, Jeremy C. Waltmunson, Kunal Patel, Daniel Frederick Penn, John Bain, Using multiple modality input to feedback context for natural language understanding ,(2011)
Christopher Vuchetich, Kudo Tsunoda, Oscar Omar Garza Santos, Relja Markovic, Andrew Robert Campbell, Matthew Eric Haigh, Stephen Latta, Kevin Geisner, Darren Bennett, Contextual tagging of recorded data ,(2010)
Henry Hyde-Thomson, David Bijl, Speech to text conversion ,(1998)
Luis Calixto Valles, Apparatus and methods for developing conversational applications ,(2003)
Mingjing Li, Jianfeng Gao, Method and apparatus for distribution-based language model adaptation ,(2001)
Ye-Yi Wang, Alejandro Acero, Ciprian Chelba, System with composite statistical and rules-based grammar model for speech recognition and natural language understanding ,(2003)
David L. Marvit, System and method for generating videoconference transcriptions ,(2010)
Miguel Cardoso de Campos, System and method for identifying the language of written text having a plurality of different length n-gram profiles ,(1998)