A Study of Machine Learning Algorithms for Recognizing Textual Entailment

This paper presents a system that uses machine learning algorithms and a combination of data sets for the task of recognizing textual entailment. The chosen features quantify lexical, syntactic and semantic level by matching between texts and hypothesis sentences. Additionally, we created a filter which uses a set of heuristics based on Named Entities to detect cases where no entailment was found. We analyzed how the different sizes of data sets and classifiers could impact on the final overall performance of the systems. We show that the system performs better than the baseline and the average of the systems from the RTE on both two and three way tasks. We concluded that evaluating using the RTE3 test set, the model learned using MLP from the RTE3 alone outperforms other models that employed different ML algorithms and additional training data from the RTE1 and RTE 2.