Dual memory network model for sentiment analysis of review text

Shen, JiaXing; Ma, Mingyu Derek; Xiang, Rong; Lu, Qin; Vallejos, Elvira Perez; Xu, Ge; Long, Yunfei; Huang, Chu-Ren

Abstract

In sentiment analysis of product reviews, both user and product information are proven to be useful. Current works handle user profile and product information in a unified model which may not be able to learn salient features of users and products effectively. In this work, we propose a dual user and product memory network (DUPMN) model to learn user profiles and product information for reviews classification using separate memory networks. Then, the two representations are used jointly for sentiment analysis. The use of separate models aims to capture user profiles and product information more effectively. Comparing with state-of-the-art unified prediction models, evaluations on three benchmark datasets (IMDB, Yelp13, and Yelp14) show that our dual learning model gives performance gain of 0.6%, 1.2%, and 0.9%, respectively. The improvements are also deemed very significant measured by p-values.

Citation

Shen, J., Ma, M. D., Xiang, R., Lu, Q., Vallejos, E. P., Xu, G., …Huang, C.-R. (2019). Dual memory network model for sentiment analysis of review text. Knowledge-Based Systems, 188, Article 105004. https://doi.org/10.1016/j.knosys.2019.105004

• Journal Article Type: Article
• Acceptance Date: Aug 27, 2019
• Online Publication Date: Sep 6, 2019
• Publication Date: Sep 6, 2019
• Deposit Date: Oct 1, 2019
• Publicly Available Date: Sep 7, 2020
• Journal: Knowledge-Based Systems
• Print ISSN: 0950-7051
• Electronic ISSN: 1872-7409
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 188
• Article Number: 105004
• DOI: https://doi.org/10.1016/j.knosys.2019.105004
• Keywords: Software; Information Systems and Management; Management Information Systems; Artificial Intelligence
• Public URL: https://nottingham-repository.worktribe.com/output/2647374
• Publisher URL: https://www.sciencedirect.com/science/article/pii/S0950705119304198
• Contract Date: Oct 1, 2019