Chinese Complex Question Analysis Based On Event Extraction
Question analysis is one of the important components in Question-Answering system. Since the existence of events, complex question analysis becomes extremely difficult Using semantic role labeling based on CRFs, event extraction provides effective support to formal description for semantic information of the complex questions. The semantic model built according to this method improves the accuracy of the information retrieval and answer extraction. Experimental results show the effectiveness of this method.
component complex question event conditional random fields
Liu Xiaoming liu li
School of computer Beijing Institute of Technology Beijing, China
国际会议
太原
英文
667-670
2010-10-22（万方平台首次上网日期，不代表论文的发表时间）