On the Recursive Neural Networks for Relation Extraction and Entity Recognition

The models are inspired by several recent works in deep learning for natural language and avoid using any external features, so as to investigate the power of representation instead of feature engineering.

Abstract

Recently there has been a surge of interest in neural architectures for complex structured learning tasks. Along this track, we are ad-dressing the supervised task of relation extrac-tion and named-entity recognition via recur-sive neural structures and deep unsupervised feature learning. Our models are inspired by several recent works in deep learning for nat-ural language. We have extended the pre-vious models, and evaluated them in various scenarios, for relation extraction and named-entity recognition. In the models, we avoid using any external features, so as to inves-tigate the power of representation instead of feature engineering. We implement the mod-els and proposed some more general models for future work. We will briefly review pre-vious works on deep learning and give a brief overview of recent progresses relation extrac-tion and named-entity recognition. 1