Twitter Sentiment Analysis for Covid-19 Vaccines in Turkey based on BiGRU-CNN Deep Learning Model

Abstract

Nowadays, social media platforms are the best way to express emotions. For nearly two years, the emergence of the Covid-19 new coronavirus epidemic has created unprecedented complex emotions on people in our country as well as all over the world. People's emotions became more complex after the start of vaccine studies against Covid-19. More recently, Covid-19's Delta, Omicron etc. The emergence of variants also created a great fear in the society again. In this process, people turned to social media tools to share their feelings and thoughts. Achieving sentiment analysis on Twitter is a very important and challenging task. The aim of this study is to investigate the different feelings of the Turkish people about the vaccination process by making use of the power of deep learning architectures and to provide an overview of the public's reactions to the current vaccination initiatives. In the study, Turkish tweets shared on Twitter between 16 June 2021 and 18 September 2021 were collected. People's feelings about vaccines of all kinds were assessed using TextBlob, a natural language processing (NLP) tool. Next, a new model for emotion classification was proposed. The proposed model is the BiGRU-CNN model using a single-layer Bi-directional Gateway Recurrent Unit (Bi-GRU) and Convolutional Neural Network (CNN) model with the Glove word embedding vector. The experimental results of the proposed method are promising when compared with the latest models. This work improves understanding of the public's views on COVID-19 vaccines and supports the goal of eradicating the coronavirus disease from the world.

Keywords

• Covid-19 vaccines
• Sentiment Analysis
• Deep Learning
• Glove
• CNN
• BiGRU

BiGRU-CNN Tabanlı Derin Öğrenme Modeliyle Türkiye’deki Covid-19 Aşılarına Yönelik Twitter Duygu Analizi

Abstract

Günümüzde, sosyal medya platformları duyguları ifade etmenin en iyi yoludur. Yaklaşık iki yıldır, Covid-19 yeni koronavirüs salgının ortaya çıkması tüm dünyada olduğu gibi ülkemizde de insanların üzerinde benzeri görülmemiş karmaşık duygular yarattı. Covid-19’a karşı aşı çalışmalarının başlamasından sonra insanların duyguları daha karmaşık hale geldi. Daha yakın zamanda, Covid-19’un Delta, Omicron vb. varyantlarının çıkması da toplumda yeniden büyük bir korku yarattı. İnsanlar, bu süreçte duygu ve düşüncelerini paylaşmak üzere Twitter gibi sosyal medya araçlarına yöneldi. Twitter’da duygu analizi yapmak çok önemli ve zorlu bir görevdir. Bu çalışmada amacımız, derin öğrenme mimarilerinin gücünden faydalanarak Türk halkının aşılama süreciyle ilgili farklı duygularını araştırmak ve halkın mevcut aşılama girişimlerine yönelik tepkilerine genel bir bakış sunmaktır. Çalışmada, Twitter’da 16 Haziran 2021 ve 18 Eylül 2021 arasında paylaşılan Türkçe tweetler toplanmıştır. İnsanların her türden aşılarla ilgili duyguları, doğal dil işleme (NLP) aracı olan TextBlob kullanılarak değerlendirildi. Daha sonra, duygu sınıflandırması için yeni bir model önerildi. Önerilen model, Glove kelime gömme vektörüyle tek katmanlı Çift-yönlü Geçitli Tekrarlayan Birim (Bi-GRU) ve Evrişimli Sinir Ağı (CNN) modelini kullanan BiGRU-CNN modelidir. Önerilen yöntemin deneysel sonuçları en son modellerle kıyaslandığında umut vericidir. Bu çalışma, halkın COVID-19 aşıları hakkındaki görüşlerinin anlaşılmasını geliştirmekte ve koronavirüsü dünyadan yok etme hedefini desteklemektedir.

Keywords

• Covid-19 aşı
• Duygu Analizi
• Derin Öğrenme
• Glove
• CNN
• BiGRU

References

1. Agarwal, A., Xie, B., Vovsha, I., Rambow, O. ve Passonneau, R. J. (2011, June). Sentiment analysis of twitter data. In Proceedings of the workshop on language in social media (LSM 2011) (pp. 30-38).
2. Alıcılar, H. E. ve Meltem, Ç. Ö. L. (2021). Yeni Koronavirüs Hastalığına Karşı Aşılanma Tutumu. Yeni Koronavirüs Pandemisi Sürecinde Türkiye'de Covıd-19 Aşılaması Ve Bağışıklama Hizmetlerinin Durumu, 61.
3. Ashraf, I., Hur, S. ve Park, Y. (2018). BLocate: A building identification scheme in GPS denied environments using smartphone sensors. Sensors, 18(11), 3862.
4. Aygün, İ., Kaya, B. ve Kaya, M. (2021). Aspect Based Twitter Sentiment Analysis on Vaccination and Vaccine Types in COVID-19 Pandemic with Deep Learning. IEEE Journal of Biomedical and Health Informatics.
5. Baccianella, S., Esuli, A. ve Sebastiani, F. (2010, May). Sentiwordnet 3.0: An enhanced lexical resource for sentiment analysis and opinion mining. In Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC'10).
6. Badgujar, K. C., Badgujar, V. C. ve Badgujar, S. B. (2020). Vaccine development against coronavirus (2003 to present): An overview, recent advances, current scenario, opportunities and challenges. Diabetes & Metabolic Syndrome: Clinical Research & Reviews.
7. Barkur, G. ve Vibha, G. B. K. (2020). Sentiment analysis of nationwide lockdown due to COVID 19 outbreak: Evidence from India. Asian journal of psychiatry, 51, 102089.
8. Bradley, M. M. ve Lang, P. J. (1999). Affective norms for English words (ANEW): Instruction manual and affective ratings (Vol. 30, No. 1, pp. 25-36). Technical report C-1, the center for research in psychophysiology, University of Florida.

9. Cabanillas, B. ve Novak, N. (2021). Allergy to COVID-19 vaccines: a current update. Allergology International, 70(3), 313-318.
10. Cambria, E., Havasi, C. ve Hussain, A. (2012, May). Senticnet 2: A semantic and affective resource for opinion mining and sentiment analysis. In Twenty-Fifth international FLAIRS conference.
11. Chakraborty, K., Bhatia, S., Bhattacharyya, S., Platos, J., Bag, R. ve Hassanien, A. E. (2020). Sentiment Analysis of COVID-19 tweets by Deep Learning Classifiers—A study to show how popularity is affecting accuracy in social media. Applied Soft Computing.
12. Cho, K., Van Merriënboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H. ve Bengio, Y. (2014). Learning phrase representations using RNN encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078.
13. Cortes Corinna, Vapnik Vladimir. Support-vector networks. Machine learning. 1995; 20(3):273–297.
14. Depoux, A., Martin, S. ve Karafillakis, E. (2020). Raman Preet, Annelies Wilder-Smith, and Heidi Larson. The pandemic of social media panic travels faster than the covid-19 outbreak.
15. Demir, F. (2021). L1-Norm DVM ve Ki-Kare Tabanlı Öznitelik Seçme Algoritmaları ile Parkinson Hastalığının Konuşma Sinyalleri Üzerinden Saptanması. International Journal of Pure and Applied Sciences, 7(1), 32-40.
16. Duncan, B. ve Zhang, Y. (2015, July). Neural networks for sentiment analysis on Twitter. In 2015 IEEE 14th International Conference on Cognitive Informatics & Cognitive Computing (ICCI* CC) (pp. 275-278). IEEE.
17. Feng, Y. ve Zhou, W. (2020). Is working from home the new norm? an observational study based on a large geo-tagged covid-19 twitter dataset. arXiv preprint arXiv:2006.08581.
18. Go, A., Bhayani, R. ve Huang, L. (2009). Twitter sentiment classification using distant supervision. CS224N project report, Stanford, 1(12), 2009.
19. Han, J., Pei, J. ve Kamber, M. (2011). Data mining: concepts and techniques. Elsevier.
20. Hilbe, J. M. (2009). Logistic regression models. Chapman and hall/CRC.
21. Hornung, O., Dittes, S. ve Smolnik, S. (2018). When emotions go social–understanding the role of emotional intelligence in social network use.
22. https://covid19.saglik.gov.tr/ (accessed 12 march, 2022).
23. Hu, M. ve Liu, B. (2004, August). Mining and summarizing customer reviews. In Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining (pp. 168-177).
24. Hu, M. ve Liu, B. (2004, August). Mining and summarizing customer reviews. In Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining (pp. 168-177).
25. Investor fact sheet. Twitter. 2021. [29-04-2021] https://s22.q4cdn.com/826641620/files/doc_financials/2021/q1/Q1'21-Shareholder-Letter.pdf.
26. Jelodar, H., Wang, Y., Orji, R. ve Huang, S. (2020). Deep sentiment classification and topic discovery on novel coronavirus or COVID-19 online discussions: NLP using LSTM recurrent neural network approach. IEEE Journal of Biomedical and Health Informatics.
27. Joshi, R. ve Tekchandani, R. (2016, August). Comparative analysis of Twitter data using supervised classifiers. In 2016 International conference on inventive computation technologies (ICICT) (Vol. 3, pp. 1-6). IEEE.
28. Kamyab, M., Liu, G. ve Adjeisah, M. (2021). Attention-Based CNN and Bi-LSTM Model Based on TF-IDF and GloVe Word Embedding for Sentiment Analysis. Applied Sciences, 11(23), 11255.
29. Kwok, S. W. H., Vadde, S. K. ve Wang, G. (2021). Twitter Speaks: An Analysis of Australian Twitter Users' Topics and Sentiments About COVID-19 Vaccination Using Machine Learning. Journal of Medical Internet Research.
30. Li S, Wang Y, Xue J, Zhao N ve Zhu, T. The impact of COVID-19 epidemic declaration on psychological.
31. Loria, S. (2018). textblob Documentation. Release 0.15, 2, 269.
32. Lu, X. ve Zhang, H. (2021). Sentiment analysis method of network text based on improved at-bigru model. Scientific Programming, 2021.
33. Madasu, A. (2019). A Study of Feature Extraction techniques for Sentiment Analysis. arXiv preprint arXiv:1906.01573.
34. Ortiz-Sánchez, E., Velando-Soriano, A., Pradas-Hernández, L., Vargas-Román, K., Gómez-Urquiza, J. L., Cañadas-De la Fuente, G. A. ve Albendín-García, L. (2020). Analysis of the anti-vaccine movement in social networks: a systematic review. International journal of environmental research and public health, 17(15), 5394.
35. Pang, B. ve Lee, L. (2004). A sentimental education: Sentiment analysis using subjectivity summarization based on minimum cuts. arXiv preprint cs/0409058.
36. Pano, T. ve Kashef, R. (2020). A complete VADER-based sentiment analysis of bitcoin (BTC) tweets during the era of COVID-19. Big Data and Cognitive Computing, 4(4), 33.
37. Parikh, R. ve Movassate, M. (2009). Sentiment analysis of user-generated twitter updates using various classification techniques. CS224N Final Report, 118.
38. Pennebaker, J. W., Boyd, R. L., Jordan, K. ve Blackburn, K. (2015). The development and psychometric properties of LIWC2015.
39. Pennington, J., Socher, R. ve Manning, C. D. (2014, October). Glove: Global vectors for word representation. In Proceedings of the 2014 conference on empirical methods in natural language processing (EMNLP) (pp. 1532-1543).
40. Rehman, M., Tauseef, I., Aalia, B., Shah, S. H., Junaid, M. ve Haleem, K. S. (2020). Therapeutic and vaccine strategies against SARS-CoV-2: past, present and future. Future Virology, 15(7), 471-482.
41. Sohangir, S., Petty, N. ve Wang, D. (2018, January). Financial sentiment lexicon analysis. In 2018 IEEE 12th international conference on semantic computing (ICSC) (pp. 286-289). IEEE.
42. Süral, I., Griffiths, M. D., Kircaburun, K. ve Emirtekin, E. (2019). Trait emotional intelligence and problematic social media use among adults: The mediating role of social media use motives. International Journal of Mental Health and Addiction, 17(2), 3.
43. Şengür, D. (2019). KOVİD-19 Salgını Sırasında Öğrencilerin Öğrenme Alışkanlıklarının Schur Ayrıştırma Tabanlı Dalgacık Aşırı Öğrenme Makineleri ile Tahmini. International Journal of Pure and Applied Sciences, 7(1), 13-18.
44. Trapani, D. ve Curigliano, G. (2021). COVID-19 vaccines in patients with cancer. The Lancet Oncology, 22(6), 738-739.
45. We used MAXQDA 2020 (VERBI Software, 2019) for data analysis.
46. World Health Organization. Mental Health and Psychosocial Considerations during the COVID-19 Outbreak. Available online: 1 November 2020.
47. Worldometer. Available online: https://www.worldometers.info/coronavirus/?utm_campaign=homeAdvegas1? (accessed on 14 March 2022).
48. Xue, J., Chen, J., Hu, R., Chen, C., Zheng, C., Su, Y. ve Zhu, T. (2020). Twitter discussions and emotions about the COVID-19 pandemic: Machine learning approach. Journal of medical Internet research, 22(11), e20550.
49. Yan, L. C., Yoshua, B. ve Geoffrey, H. (2015). Deep learning. nature, 521(7553), 436-444.