Combination of Convolutional Neural Network and Gated Recurrent Unit for Aspect-Based Sentiment Analysis

Narisa Zhao,Hui Li,Xin Wen,Huan Gao

doi:10.1109/access.2021.3052937

Abstract

Aspect-based sentiment analysis (ABSA) aims to identify views and sentiment polarities towards a given aspect in reviews. Compared with general sentiment analysis, ABSA can provide more detailed and complete information. Recently, ABSA has become an important task for natural language understanding and has attracted considerable attention from both academic and industry fields. The sentiment polarity of a sentence is not only decided by its content but also has a relatively significant correlation with the targeted aspect. For this reason, we propose a model for aspect-based sentiment analysis which is a combination of Convolutional Neural Network (CNN) and Gated Recurrent Unit (GRU), utilizing the local features generated by CNN and the long-term dependency learned by GRU. Extensive experiments have been conducted on datasets of hotels and cars, and results show that the proposed model achieves excellent performance in terms of aspect extraction and sentiment classification. Experiments also demonstrate the great domain expansion capability of the model.

Highlights

The rapid development of e-commerce has brought about the accumulation of abundant consumer-generated reviews
We propose a combined Convolutional Neural Network (CNN) and Gated Recurrent Unit (GRU) neural network model that utilizes local features generated by CNN as the input of GRU for Aspect-based sentiment analysis (ABSA) on Chinese online review datasets of hotels and cars
We propose a deep neural network model for aspect-based sentiment analysis utilizing the combination of CNN and GRU

Summary

Introduction

The rapid development of e-commerce has brought about the accumulation of abundant consumer-generated reviews. These reviews are highly valuable in both economic and social aspects because we can extract the opinions of users by operating sentiment analysis on the product reviews. The opinions obtained can help potential consumers with purchase analysis and provide the advantages and disadvantages of the goods to providers so that they can make further improvements in their products or services. Named opinion mining [2], is an essential task in natural language processing and has attracted considerable attention both in academia and industry fields, for identifying consumer satisfaction with products and services. The work is established on the basis that there is just

Objectives

Methods

Findings

Conclusion