Affiliation:
1. Future Institute of Technology
Abstract
Abstract
Sentiment analysis is a vastly established domain for social media monitoring, feedback insights, and commercial or political campaigns, as it allows us to gain an overview of the wider public opinion on certain topics. Nowadays, different social media platforms play a crucial role in web-based sentiment analysis and emotion detection from distinct perspectives. Likewise, WhatsApp is probably the most popular messaging app, allowing users to send messages, images, audio, or videos. However, it is still highly under-explored for any type of linguistic synthesis and analysis. Like many other groups of people, students use WhatsApp for various purposes, even more since the last two years of the pandemic phase. For instance, class communication, study group communication, etc. In this paper, we present a novel approach to analyze the sentiments and emotions of students in valence, arousal, and dominance space by classifying the messages from their WhatsApp group chat. The emotional dimensions of Valence, Arousal and Dominance (VAD) can derive a person’s interest (attraction), level of activation, and perceived level of control for a particular situation from textual communication. We propose a vanilla SVM model fused with a language classifier to calculate each message's sentiment ratings. Finally, using the SVM classifier we classify the sentiment ratings concerning the degree of the VAD scale. The data were analyzed using a qualitative content analysis method. The results of the study in the form of cumulative sentiment scale and sentiment clustering in VAD space reveal that the students' WhatsApp groups were mostly used for sharing information, exchanging ideas, and discussing issues, with mostly neutral to positive sentiment viewpoints for the provided topics of discussions.
Publisher
Research Square Platform LLC
Reference37 articles.
1. Pak, A., & Paroubek, P. (2010, May). Twitter as a corpus for sentiment analysis and opinion mining. In LREc (Vol. 10, No. 2010, pp. 1320–1326).
2. Twitter sentiment analysis for large-scale data: an unsupervised approach;Pandarachalil R;Cognitive computation,2015
3. Das, S., & Kolya, A. K. (2017, November). Sense GST: Text mining & sentiment analysis of GST tweets by Naive Bayes algorithm. In 2017 third international conference on research in computational intelligence and communication networks (ICRCICN) (pp. 239–244). IEEE.
4. Sentiment classification with GST tweet data on LSTM based on polarity-popularity model;Das S;Sadhana,2020
5. Turney, P. (2002, July). Thumbs Up or Thumbs Down? Semantic Orientation Applied to Unsupervised Classification of Reviews. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics (pp. 417–424).