Recognizing Textual Entailment Using Multiple Features

摘要

Textual entailment is knowledge that may prove useful for a variety of applications dealing with inferencing over sentences described in natural language texts. This paper proposes a method to calculate the similarity between two text fragments T and H which is based on the TF-IDF algorithm. Similarity calculation, which contains synonymous degree and particular degree, is an important part in our system. System designed to recognize textual entailment typically employ lexical information. We analyze the experimental results and prospect for follow-up work. The evaluation results show that our method is effective for RTE task.