Abstract:In the circumstance of social big data, sentiment analysis is attracting increasing attention for its capacity in understanding individuals' attitudes and feelings. Traditional sentiment analysis methods focus on single modality and become ineffective as enormous data are emerging on the social websites with multiple manifestations. In this article, multimodal learning approaches are proposed to capture the relations between image and text, which only stay at the region level and ignore the fact that the channels are also closely correlated with the semantic information. In addition, social images in the social platforms are closely connected by various types of relations, which are also conducice to sentiment classification but neglected by most existing works. In this article, we propose an attention-based heterogeneous relational model to improve the multimodal sentiment analysis performance by incorporating rich social information. Specifically, we propose a progressive dual attention module to capture the correlations between image and text, and then learn the joint image-text representation from the perspective of content information. A channel attention schema is proposed here to highlight semantically rich image channels and a region attention schema is further designed to highlight the emotional regions based on the attended channels. After that, we construct a heterogeneous relation network and extend graph convolutional network to aggregate the content information from social contexts as complements to learn high-quality representations of social images. Our proposal is thoroughly evaluated on two benchmark datasets, and experimental results demonstrate the superiority of the proposed model.

Sentiment Analysis on the People’s Daily

Reflections on Sentiment/Opinion Analysis

Topic Sentiment Analysis in Chinese News

Research on Sentiment Dictionary Based on Sentiment Analysis in News Domain.

Pretrain Deep Models by Distant Supervision for Weibo Sentiment Analysis

The Public Sentiment Analysis of Double Reduction Policy on Weibo Platform

Exploiting Discourse Relations for Sentiment Analysis.

Chinese Implicit Sentiment Analysis Based on Hierarchical Knowledge Enhancement and Multi-Pooling

Learning From User Social Relation For Document Sentiment Classification

Sentiment Analysis With The Exploration Of Overall Opinion Sentences

Hierarchical Structure Based Hybrid Approach to Sentiment Analysis of Chinese Micro Blog and Its Feature Extraction

A Hybrid Method of Sentiment Analysis and Machine Learning Algorithm for the U.S. Presidential Election Forecasting

Cooperative Hybrid Semi-Supervised Learning For Text Sentiment Classification

A Novel Sentiment Polarity Detection Framework for Chinese

Constructing Financial Sentimental Factors in Chinese Market Using Natural Language Processing

Using Bilingual Knowledge and Ensemble Techniques for Unsupervised Chinese Sentiment Analysis.

Interpreting the Public Sentiment Variations on Twitter

Stock Market Prediction Exploiting Microblog Sentiment Analysis.

Social Image Sentiment Analysis by Exploiting Multimodal Content and Heterogeneous Relations

Sentiment Classification in Chinese Microblogs: Lexicon-based and Learning-based Approaches

User-level Sentiment Analysis Incorporating Social Networks