Recognizing Textual Entailment (RTE) is a PASCAL/TAC task in which two text fragments are processed by system to determine whether the meaning of hypothesis is entailed from another text or not. In this paper, we proposed a textual entailment system using a statistical approach that integrates syntactic and semantic techniques for Recognizing Inference in Text (RITE) using the NTCIR-9 RITE task and make a comparison between semantic and syntactic features based on their differences. We thoroughly evaluate our approach using subtasks of the NTCIR-9 RITE. As a result, our system achieved 73.28% accuracy on the Chinese Binary-Class (BC) subtask with NTCIR-9 RITE. Thorough experiments with the text fragments provided by the NTCIR-9 RITE task show that the proposed approach can significantly improve system accuracy.
關聯:
Proceedings of the IEEE International Conference on Information Reuse and Integration (IEEE IRI 2012), pp.59-64
