InterSentiment: combining deep neural models on interaction and sentiment for review rating prediction
- PDF / 1,904,496 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 14 Downloads / 186 Views
ORIGINAL ARTICLE
InterSentiment: combining deep neural models on interaction and sentiment for review rating prediction Shi Feng1 · Kaisong Song2 · Daling Wang1 · Wei Gao3 · Yifei Zhang1 Received: 16 September 2019 / Accepted: 1 August 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Review rating prediction is commonly approached from the perspective of either Collaborative Filtering (CF) or Sentiment Classification (SC). CF-based approach usually resorts to matrix factorization based on user–item interaction, and does not fully utilize the valuable review text features. In contrast, SC-based approach is focused on mining review content, but can just incorporate some user- and product-level features, and fails to capture sufficient interactions between them represented typically in a sparse matrix as CF can do. In this paper, we propose a novel, extensible review rating prediction model called InterSentiment by bridging the user-product interaction model and the sentiment model based on deep learning. InterSentiment is a specific instance of our proposed Deep Learning based Collaborative Filtering framework. The proposed model aims to learn the high-level representations combining user-product interaction and review sentiment, and jointly project them into the rating scores. Results of experiments conducted on IMDB and two Yelp datasets demonstrate clear advantage of our proposed approach over strong baseline methods. Keywords Review rating prediction · Deep neural networks · Matrix factorization · Sentiment analysis · User–product interaction
1 Introduction Review rating prediction is an important sentiment analysis task which aims to detect users’ sentiment intensity towards target products from vast amount of subjective reviews on
* Shi Feng [email protected] Kaisong Song kaisong.sks@alibaba‑inc.com Daling Wang [email protected] Wei Gao [email protected] Yifei Zhang [email protected] 1
School of Computer Science and Engineering, Northeastern University, No. 195 Chuangxin Road, Hunnan District, Shenyang, China
2
Alibaba Group Hangzhou, Hangzhou, China
3
Victoria University of Wellington, Wellington, New Zealand
online websites (e.g., 1–5 stars in Yelp, or 1–10 stars in IMDB). Early research approaches to the task from either the angle of Sentiment Classification (SC) or that of Collaborative Filtering (CF). SC-based models primarily follow Pang and Lee [18] by concentrating text mining and regard the problem as a Single-Label Multi-Class (SLMC) classification task. Most of studies in this approach rely on handcrafted features and/or sentiment lexicons for achieving effective learning performance, which however is biased and labor intensive [5, 8]. Recently, neural network based models have achieved promising SC results. These models have strong representation learning capacity that can capture and organize discriminative features automatically extracted from data. Some of such studies have noticed the importance of user and product elements on
Data Loading...
