Electrotechnical and Computer Engineering
Vol. 38 No. 07 (2023): Proceedings of Faculty of Technical Sciences
ANNOTATION AND SENTIMENT ANALYSIS OF TWEETS RELATED TO THE POLITICAL SCENE OF THE REPUBLIC OF SERBIA
Abstract
This paper describes tweet sentiment analysis and the development of a tool for placing tweet authors on a political map. The purpose of this system is to assist its users in political elections. BERTić model is used for tweet sentiment analysis. Sentiment and tweet’s theme are used as inputs for SVM and Random-forest, which are used for user polarization.
References
[1] J. D. M.-W. C. K. L. K. Toutanova, „BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding“.
[2] B. G. J. R. A. F. F. M. Michael D. Conover, Predicting the Political Alignment of Twitter Users.
[3] E. L.-G. Jhon Adrian Ceron-Guzman, A Sentiment Analysis System of Spanish Tweets and Its Application in Colombia 2014 Presidential Election.
[4] D. L. Nikola Ljubešić, „BERTić - The Transformer Language Model for Bosnian, Croatian, Montenegrin and Serbian“.
[5] „EMBEDDIA/bertic-tweetsentimen,“ [На мрежи]. Available: https://huggingface.co/EMBEDDIA/bertic-tweetsentimen.
[6] „SrbAi,“ [На мрежи]. Available: https://github.com/Serbian-AI-Society/SrbAI.
[7] „Scikit-learn,“ [На мрежи]. Available: https://scikit-learn.org/stable/.
[8] М. Кнежевић, „Позиционирање корисника друштвене мреже Twitter на мапи политичког спектра помоћу корисничких твитова“.