Recognize Textual Entailment by the Lexical and Semantic Matching
Recognizing Textual Entailment (RTE) task is defined as recognizing, given two text fragments, whether the meaning of one text can be inferred from the other. This paper adopts the word overlap technique at first for detection the entailment relation, and then the semantic information is also combined for this task by the use of WordNet which is a structured lexical knowledge base. The part-of-speech and named entity can give useful indication for entailment recognition and we give additional attention to them. The experiments on the RTE test set get the encouraged results and we give the result analysis and we will mine the deeper semantic information in the future work.
textual entailment lexical matching semantic matching
Yongping Du Changqing Yao Jiangli Liu
Institute of Computer Science Beijing University of Technology Beijing, China Institute of Scientific and Technical Information of China Beijing, China
国际会议
太原
英文
500-504
2010-10-22（万方平台首次上网日期，不代表论文的发表时间）