ABSTRACT
To exactly classify sentiments of microblog reviews with emojis in microblog social networks, this paper first proposes an emoji vectorisation method to achieve emoji vectors. Then, an emoji-text integrated bidirectional LSTM (ET-BiLSTM) model for sentiment analysis is proposed. In this model, review text-based sentence representations are extracted by a bidirectional LSTM network. Emoji-based auxiliary representations are obtained by a new attention mechanism. The two representations are further integrated into final review representation vectors. Finally, experimental results indicate that the proposed ET-BiLSTM model improves the performance of sentiment classification evaluated by macro-P, macro-R and macro-F1 scores in microblog social networks.
Acknowledgments
This work is partially supported by the National Natural Science Foundation (Nos. 61802316, 61872298 and 61902324), the Chunhui Plan Cooperation and Research Project, Ministry of Education of China (Nos. Z2015109, Z2015100), the “Young Scholars Reserve Talents” program of Xihua University, the Science and Technology Department of Sichuan Province (Nos. 22ZDYF3157, 2021YFQ0008), the Key Scientific Research Fund of Xihua University (No. z1422615), and the Discipline construction project of Guangdong Medical University (No. 4SG21018G).
Disclosure statement
No potential conflict of interest was reported by the author(s).