DLSITE-2: Semantic Similarity Based on Syntactic Dependency Trees Applied to Textual Entailment
Daniel Micol, Óscar Ferrández, Rafael Muñoz, Manuel Palomar · 2007
In this paper we attempt to deduce tex-tual entailment based on syntactic depen-dency trees of a given text-hypothesis pair. The goals of this project are to provide an accurate and fast system, which we have called DLSITE-2, that can be applied in software systems that require a near-real-time interaction with the user. To accom-plish this we use MINIPAR to parse the phrases and construct their correspond-ing trees. Later on we apply syntactic-based techniques to calculate the seman-tic similarity between text and hypothe-sis. To measure our method’s precision we used the test text corpus set from Second PASCAL Recognising Textual Entailment Challenge (RTE-2), obtaining an accuracy rate of 60.75%. 1