作者: Alex Acero , Ye-Yi Wang
DOI:
关键词:
摘要: SGStudio is a grammar authoring tool that eases semantic development. It capable of integrating different information sources and learning from annotated examples to induct CFG rules. In this paper, we investigate modification its underlying model by replacing rules with n-gram statistical models. The new composite HMM CFG. advantages the include built-in robust feature scalability an classifier when understanding does not involve slot filling. We devised decoder for model. Preliminary results show achieved 32% error reduction in high resolution understanding.