Textual Inference with Tree-structured LSTMs1

Textual Inference is a research trend in Natural Language Processing (NLP) that has recently received a lot of attention by the scientific community. Textual Entailment (TE) is a specific task in Textual Inference that aims at determining whether a hypothesis is entailed by a text. Usually tackled by machine learning techniques employing features which represent similarity between texts, the recent availability of more training data presupposes that Neural Networks that are able to learn latent feature from data for generalized prediction could be employed. This paper employs the Child-Sum TreeLSTM for solving the challenging problem of textual entailment. Our approach is simple and able to generalize well without excessive parameter optimization. Evaluation done on SNLI, SICK and other TE datasets shows the competitiveness of our approach.