作者: Mazin G. Rahim , Srinivas Bangalore , Narendra K. Gupta
DOI:
关键词: Dialog act 、 Action (philosophy) 、 Natural language processing 、 Spoken dialog 、 Speech recognition 、 Spoken language 、 Artificial intelligence 、 Computer science 、 Process (computing) 、 Object (computer science) 、 Domain (software engineering) 、 Dialog box
摘要: A system and method are disclosed that improve automatic speech recognition in a spoken dialog system. The comprises partitioning recognizer output into self-contained clauses, identifying act each of the qualifying acts by current domain object and/or action, determining whether further qualification is possible for action. If possible, then another action associated with reassigning as recursively new object. This process continues until nothing left to qualify.