Association Extraction and Recognition of Multiple Emotion Expressed in Social Texts
Zou J, Wu S, Yang Z, Chen C, Sun Y, Jiang M, Huang Y (2022)
Publication Type: Conference contribution
Publication year: 2022
Journal
Publisher: Springer Science and Business Media Deutschland GmbH
Book Volume: 13338 LNCS
Pages Range: 420-430
Conference Proceedings Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Event location: Qinghai
ISBN: 9783031067938
DOI: 10.1007/978-3-031-06794-5_34
Abstract
Detecting the sentiment people present in social media such as tweets is important for politics, commerce, education and so on. The task of multiple emotion recognition in texts is to predict a set of emotion labels that expressed in sentences. There are still some shortcomings in the current works: 1) the dependencies among emotions are not well modeled due to the complex combinatorial features of them, 2) the semantics of emotion labels as well as the semantic correlations between emotion labels and sentences are not fully considered. In this paper, in the purpose of capturing the dependencies between emotions, we propose a new method by using Graph Convolutional Network (GCN) based on a label co-occurrence matrix building from the dataset, and a Convolutional Neural Network (CNN) is used to capture the syntactic and semantic information in the sentences through different convolutional filters, the outputs of GCN and CNN are multiplied together to fuse their features as the last output. Experiments on SemEval2018 Task1: E-c multi-label emotion recognition problem show that metrics have been significantly improved, and our approach obviously obtains the dependencies among emotions described by Pointwise Mutual Information (PMI) which measures the correlations between emotions both in the true test labels and predicted labels.
Authors with CRIS profile
Involved external institutions
How to cite
APA:
Zou, J., Wu, S., Yang, Z., Chen, C., Sun, Y., Jiang, M., & Huang, Y. (2022). Association Extraction and Recognition of Multiple Emotion Expressed in Social Texts. In Xingming Sun, Xiaorui Zhang, Zhihua Xia, Elisa Bertino (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 420-430). Qinghai, CN: Springer Science and Business Media Deutschland GmbH.
MLA:
Zou, Jiajun, et al. "Association Extraction and Recognition of Multiple Emotion Expressed in Social Texts." Proceedings of the 8th International Conference on Artificial Intelligence and Security, ICAIS 2022, Qinghai Ed. Xingming Sun, Xiaorui Zhang, Zhihua Xia, Elisa Bertino, Springer Science and Business Media Deutschland GmbH, 2022. 420-430.
BibTeX: Download