WUST at NTCIR-10 RITE-2 Task: Multiple Feature Approach to Chinese Textual Entailment.
Maofu Liu, Yue Wang, Yan Li, Huijun Hu · 2013
Conclusions We construct the classification model based on support vector machine to recognize semantic inference in Chinese text pair using multiple features, including statistical, syntactic and lexical semantic ones. Through further analysis, we find that in our system, we use the same features in BC and MC subtasks, but the characteristics of the BC and MC subtasks should be different, which may cause the dissatisfaction of BC result. Moreover, we mostly consider statistical features in our system, if we add some rule features, the accuracy may be significantly improved. Euclidean distance