End-to-end adversarial memory network for cross-domain sentiment classification

Abstract

Domain adaptation tasks such as cross-domain sentiment classification have raised much attention in recent years. Due to the domain discrepancy, a sentiment classifier trained in a source domain may not work well when directly applied to a target domain. Traditional methods need to manually select pivots, which behave in the same way for discriminative learning in both domains. Recently, deep learning methods have been proposed to learn a representation shared by domains. However, they lack the interpretability to directly identify the pivots. To address the problem, we introduce an endto-end Adversarial Memory Network (AMN) for cross-domain sentiment classification. Unlike existing methods, the proposed AMN can automatically capture the pivots using an attention mechanism. Our framework consists of two parametershared memory networks with one for sentiment classification and the other for domain classification. The two networks are jointly trained so that the selected features minimize the sentiment classification error and at the same time make the domain classifier indiscriminative between the representations from the source or target domains. Moreover, unlike deep learning methods that cannot tell which words are the pivots, AMN can offer a direct visualization of them. Experiments on the Amazon review dataset demonstrate that AMN can significantly outperform state-of-the-art methods.