wCaWESA - weighted Capitalized Word Enhanced Sentiment Analysis Model
Abstract
Sentiment analysis is a crucial task in natural language processing (NLP) that aims to determine the emotional tone of text. Traditional sentiment analysis models often overlook the impact of word formatting, particularly capitalization, on sentiment intensity. In informal communication, such as social media platforms, users frequently employ all-capitalized words to emphasize their emotions, signaling a heightened sentiment. This study introduces a novel model, weighted Capitalized Word Enhanced Sentiment Analysis (wCaWESA), which addresses this gap by explicitly accounting for the sentiment strength of capitalized words. The model assigns specific weights to capitalized words based on their frequency and context, enhancing the overall accuracy of sentiment analysis. By incorporating this additional layer of sentiment intensity, wCaWESA provides a more nuanced understanding of the emotional content in text, particularly in environments where informal and emphatic language is prevalent. The proposed model demonstrates significant improvements over existing approaches, making it a valuable tool for applications ranging from social media monitoring to opinion mining.
References
2. Hutto, C. J., & Gilbert, E. (2014). VADER: A Parsimonious Rule-based Model for Sentiment Analysis of Social Media Text. Proceedings of the Eighth International Conference on Weblogs and Social Media (ICWSM-14).
3. Irina, P., & Phoey, L. T. (2018). Value of Expressions behind the Letter Capitalization in Product Reviews. 7th International Conference on Software and Computer Applications. New york.
4. Kim, Y. (2014). Convolutional neural networks for sentence classification. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), 1746-1751.
5. Kiritchenko, S. Z. (2014). Sentiment analysis of short informal texts. Journal of Artificial Intelligence Research, 50, 723-762.
6. Kiritchenko, S. Z. (2018). Sentiment analysis of tweets: The role of capital letters. Journal of Artificial Intelligence Research, 62, 493-514.
7. Kumar, R., & Vadlamani, R. (2015). A survey on opinion mining and sentiment analysis: Tasks, approaches and applications. Knowledge-Based Systems, 89, 14-46.
8. Lin, T. &. (2016). Emphasis Detection for Expressive Text. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing , 1532-1541.
9. Liu, B. (2012). Sentiment Analysis and Opinion Mining. Synthesis Lectures on Human Language Technologies, 5(1), 1-167.
10. Liu, Y. O. (2019). RoBERTa: A Robustly Optimized BERT Pretraining Approach. arXiv preprint arXiv:1907.11692.
11. McCulloch, G. (2019, July 23). The Meaning of All Caps—in Texting and in Life. (Wired) Retrieved from https://www.wired.com/story/all-caps-because-internet-gretchen-mcculloch/
12. Pang, B. &. (2008). Opinion Mining and Sentiment Analysis. Foundations and Trends in Information Retrieval, 2((1-2)), 1-135.
13. Potts, C. (2011). On the negativity of negation. In Proceedings of the 2011 Conference on Empirical Methods in Natural Language Processing, 1247-1252.
14. Thelwall, M. &. (2013). Topic-based sentiment analysis for the social web: The role of mood and issue-related words. Journal of the Association for Information Science and Technology, 64(8), 1608-1620.
15. Victor Sanh, L. D. (2019). DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter. https://arxiv.org/abs/1910.01108.
16. Wang, Y., Huang, M., & Zhao, L. (2018). Attention-based LSTM for Aspect-level Sentiment Classification. Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, 606-615.
17. Wankhade, M., Rao, A., & Kulkarni, C. (2022). A survey on sentiment analysis methods, applications, and challenges. Artif Intell Rev, 55, 5731–5780.
