An Entropy-Based Method with a New Benchmark Dataset for Chinese Textual Affective Structure Analysis

Xiong, Shufeng; Fan, Xiaobo; Batra, Vishwash; Zeng, Yiming; Zhang, Guipei; Xi, Lei; Liu, Hebing; Shi, Lei

Abstract

Affective understanding of language is an important research focus in artificial intelligence. The large-scale annotated datasets of Chinese textual affective structure (CTAS) are the foundation for subsequent higher-level analysis of documents. However, there are very few published datasets for CTAS. This paper introduces a new benchmark dataset for the task of CTAS to promote development in this research direction. Specifically, our benchmark is a CTAS dataset with the following advantages: (a) it is Weibo-based, which is the most popular Chinese social media platform used by the public to express their opinions; (b) it includes the most comprehensive affective structure labels at present; and (c) we propose a maximum entropy Markov model that incorporates neural network features and experimentally demonstrate that it outperforms the two baseline models.

Citation

Xiong, S., Fan, X., Batra, V., Zeng, Y., Zhang, G., Xi, L., …Shi, L. (2023). An Entropy-Based Method with a New Benchmark Dataset for Chinese Textual Affective Structure Analysis. Entropy, 25(5), Article 794. https://doi.org/10.3390/e25050794

• Journal Article Type: Article
• Acceptance Date: May 11, 2023
• Publication Date: May 13, 2023
• Deposit Date: Jul 4, 2023
• Journal: ENTROPY
• Publisher: MDPI
• Peer Reviewed: Peer Reviewed
• Volume: 25
• Issue: 5
• Article Number: 794
• DOI: https://doi.org/10.3390/e25050794
• Keywords: affective structure; corpus annotation; Chinese benchmark datasets; affective computing
• Publisher URL: https://www.mdpi.com/1099-4300/25/5/794