UAIC Participation at AVE 2007

Abstract

Textual Entailment Recognition (RTE) is a recently proposed task, aiming at capturing the means through which textual inferences can be made. Moreover, using such a module is meant to contribute to the increase in performance of many NLP applications, such as Summarization, Information Retrieval or Question Answering, both for answer ranking as well as for answer validation. This article presents the manner in which we used the TE system built for the RTE3 competition this year for the AVE exercise. We describe the steps followed in building the patterns for question transformation, the generation of the corresponding hypotheses and finally for answer ranking. We conclude by presenting an overview of the performance obtained by this approach and a critical analysis of the errors obtained. © 2008 Springer-Verlag Berlin Heidelberg.