Abstract

Online forums have become crucial platforms for making decisions and expressing emotions [3]. The purpose of the study is to find out how users’ decisions during online debates are influenced by their emotions, particularly happiness, anger, sadness, and enthusiasm [24]. While many studies focus on examining relationships between emotions and specific behaviors, this study aims to investigate if emotions cause those behaviors [22]. In order to accomplish this goal, psychological understanding of how people respond to particular circumstances is integrated with contemporary computational techniques [11]. Using textual data collected from online discussion forums, a deep learning transformer network will identify emotions [7]. After that, emotional expressions will be examined in terms of their dynamics and impact on user reactions. Reply rates, agreement, and interaction are all decisions made by users. The study’s integration of deep learning techniques with a causality analysis model to determine whether there is a relationship between certain user actions and emotions is one of its main achievements [28]. The models Logistic Regression, SVM, Random Forest, LSTM, and BERT will be evaluated according to their F1-score, accuracy, precision, and recall [20]. The findings demonstrate that in terms of comprehending emotional context and forecasting user behavior, transformer-based models outperform conventional techniques [17]. More significantly, the results imply that decisions made in online settings are heavily influenced by emotions [16]. All things considered, this study offers a coherent paradigm that links psychology and computer science, providing greater understanding of human behavior in online communities. Additionally, it aids in the creation of online systems that are better able to comprehend and react to human emotions.

References

• David et al. Blei. Latent dirichlet allocation. JMLR, 2003.
• Johan et al. Bollen. Twitter mood predicts stock market. Journal of Computational Science, 2011.
• Danah Boyd and Nicole Ellison. Social network sites. JCMC, 2007.
• Erik et al. Cambria. Senticnet. In AAAI, 2014.
• Dorottya et al. Demszky. Goemotions: Fine-grained emotion dataset. In ACL, 2020.
• Jacob Devlin. Bert fine-tuning, 2019.
• Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of NAACL, 2019.
• Paul Ekman. Emotions Revealed. Times Books, 2003.
• Emilio Ferrara and Zeyao Yang. Emotional contagion in social media. PLoS ONE, 2015.
• Alec et al. Go. Twitter sentiment classification using distant supervision, 2009.
• Yoav Goldberg. A primer on neural network models for nlp. JAIR, 2016.
• Sepp Hochreiter and J¨urgen Schmidhuber. Long short-term memory. Neural Computation, 9(8):1735–1780, 1997.
• Guido Imbens and Donald Rubin. Causal Inference. Cambridge University Press, 2015.
• Jan et al. Kietzmann. Social media? get serious! Business Horizons, 2011.
• Yoon Kim. Cnn for sentence classification. In EMNLP, 2014.
• Adam et al. Kramer. Experimental evidence of emotional contagion. PNAS, 2014.
• Yinhan et al. Liu. Roberta: A robustly optimized bert approach. arXiv preprint arXiv:1907.11692, 2019.
• Tomas et al. Mikolov. Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781, 2013.
• Saif Mohammad and Peter Turney.Word-emotion association lexicon. Computational Intelligence, 2013.
• Bo Pang and Lillian Lee. Opinion mining and sentiment analysis. Foundations and Trends in Information Retrieval, 2008.
• Bo et al. Pang. Sentiment classification using ml. In EMNLP, 2002.
• Judea Pearl. Causality. Cambridge University Press, 2009.
• James W. et al. Pennebaker. Linguistic Inquiry and Word Count: LIWC 2015. Pennebaker Conglomerates, 2015.
• Robert Plutchik. The nature of emotions. American Scientist, 2001.
• Alec et al. Radford. Language models are unsupervised multitask learners. 2019.
• Donald Rubin. Estimating causal effects. Journal of Educational Psychology, 1974.
• James A. Russell. A circumplex model of affect. Journal of Personality and Social Psychology, 1980.
• Uri et al. Shalit. Estimating treatment effect. In ICML, 2017.
• Richard et al. Socher. Recursive deep models. In EMNLP, 2013.
• Duyu et al. Tang. Sentiment-specific word embedding. In ACL, 2014.
• Ashish et al. Vaswani. Attention is all you need. In NeurIPS, 2017.
• Tom et al. Young. Recent trends in deep learning nlp. IEEE Computational Intelligence Magazine, 2018.
• Ye et al. Zhang. Deep learning for sentiment analysis. IEEE Intelligent Systems, 2018.