Deep Learning-Based Sentiment Analysis on Education During the COVID-19 Pandemic

Öz

The global COVID-19 pandemic in 2020 has led to catastrophic economic and social disruption. The pandemic has affected almost every aspect of our lives, including health, food, business organizations, and education. An essential shift in the higher education field has been occurred with the digitalization of instruction. In attempt to combat the pandemic, several higher education institutions throughout the world have begun to offer undergraduate and graduate courses online, either asynchronously or synchronously. During this period, people make considerable use of social media to gain news, information, social connections, and support. As a result, the immense quantity of electronic text documents has been shared on the Web related to COVID-19. In this paper, we present a deep learning-based sentiment analysis approach to analyze the impact of COVID-19 pandemic on the higher education. In this regard, the predictive performance of conventional machine learning algorithms (support vector machines, naïve bayes, logistic regression, and random forest) and deep neural networks (convolutional neural network, recurrent neural network, long short-term memory, and gated recurrent unit) are compared to each other. In addition, the empirical results obtained by the bidirectional encoder representations from transformers (BERT) have been evaluated. The comprehensive empirical results with different text representation models and classification algorithms indicate that deep neural networks can yield promising results for the task of analyzing the impact of COVID-19 related text documents on the higher education.

Anahtar Kelimeler

• Deep Learning
• Sentiment Analysis
• Text Mining
• COVID-19
• Higher Education

COVID-19 Pandemi Döneminde Eğitimde Derin Öğrenmeye Dayalı Duygu Analizi

Öz

Keywords: Deep Learning, Sentiment Analysis, Text Mining, COVID-19, Higher Education Öz 2020 yılında küresel COVID-19 pandemisi, ciddi ekonomik ve toplumsal kesintilere yol açtı. Pandemi sağlık, gıda, iş organizasyonları ve eğitim dahil olmak üzere hayatımızın neredeyse her alanını etkiledi. Eğitimin dijitalleştirilmesi ile birlikte yükseköğretim alanında önemli bir değişiklik yaşanmıştır. Pandemi ile mücadele amacıyla, dünya çapında birçok yükseköğretim kurumu, eş zamanlı veya eş zamansız olarak lisans ve lisansüstü derslerini çevrimiçi olarak sunmaya başlamıştır. Bu süre zarfında insanlar haber, bilgi, destek almak için ve sosyal bağlantılar kurmak için sosyal medyadan ciddi ölçüde yararlanmaktadırlar. Bu sayede, COVID-19 ile ilgili olarak Web'de çok miktarda elektronik metin belgesi paylaşılmıştır. Bu makalede, COVID-19 salgınının yüksek öğrenim üzerindeki etkisini analiz etmek için derin öğrenime dayalı bir duygu analizi yaklaşımı sunuyoruz. Bu bağlamda, geleneksel makine öğrenimi algoritmalarının (vektör destek makineleri, naive bayes, lojistik regresyon ve rastgele orman) ve derin sinir ağlarının (evrişimli sinir ağı, tekrarlı sinir ağı, uzun süreli bellek ve gated tekrarlı birim) performansları karşılaştırılmıştır. Buna ek olarak, transformerlardan gelen çift yönlü enkoder gösterimleri (BERT) tarafından elde edilen ampirik sonuçlar da değerlendirilmiştir. Farklı metin gösterim modelleri ve sınıflandırma algoritmalarına sahip kapsamlı ampirik sonuçlar, derin sinir ağlarının COVID-19 ile ilgili metin belgelerinin yüksek eğitim üzerindeki etkisini analiz etme görevi için umut verici sonuçlar verebileceğini göstermektedir.

Anahtar Kelimeler

• Derin Öğrenme
• Duygu Analizi
• Metin Madenciliği
• COVID-19
• Yüksek Öğrenim

Kaynakça

1. Chawla, S., Mittal, M., Chawla, M., and Chawla, and Goyal, L.M. 2020. Corona virus - SARS-CoV-2: An insight to another way of natural disaster, EAI Endorsed Trans. Pervasive Health Technol, Vol. 6, pp. 25-33. DOI: 10.4108/eai.28-5-2020 164823
2. Wang, L.L., and Lo, K. 2021. Text mining approaches for dealing with the rapidly expanding literature on COVID-19, Brief. Bioinform, Vol. 22, pp. 781–799. DOI: 10.1093/bib/bbaa296
3. Gajewski, N.K, Peterson, A.E., Chitale, R.A., and Pavlin, J.A. 2014. A review of evaluations of electronic event-based biosurveillance systems, PLoS One, Vol. 9, DOI: 10.1371/journal.pone.0111222
4. Bismala, L., and Manurung, Y.M. 2021. Student satisfaction in e-learning along the COVID-19 pandemic with importance performance analysis, Int. J. Eval. Res. Educ. (IJERE), Vol. 10, DOI: 10.11591/ijere.v10i3.21467
5. Daniel, S.J. 2020. Education and the COVID-19 pandemic, Prospects (Paris), Vol. 49, pp. 1–6. DOI: 10.1007/S11125-020-09464-3
6. Bilecen, B. 2020. Commentary: COVID‐19 pandemic and higher education: International mobility and students’ social protection, International Migration, Vol. 58, pp. 263–266. DOI: 10.1111/imig.12749
7. Onan, A., Korukoğlu, S., and Bulut, H. 2016 Ensemble of keyword extraction methods and classifiers in text classification, Expert Syst. Appl., Vol. 57, pp. 232–247. DOI: 10.1016/J.ESWA.2016.03.045
8. Onan, A. 2016. Classifier and feature set ensembles for web page classification,J. Inf. Sci., Vol. 42, pp. 150–165. DOI: 10.1177/0165551515591724

9. Onan, A., Korukoğlu, S., and Bulut, H. 2016. A multiobjective weighted voting ensemble classifier based on differential evolution algorithm for text sentiment classification, Expert Systems With Applications, Vol. 62, pp. 1–16. DOI: 10.1016/j.eswa.2016.06.005
10. Onan, A., and Korukoğlu, S. 2017. A feature selection model based on genetic rank aggregation for text sentiment classification,Journal of Information Science, Vol. 43, pp. 25–38. DOI: 10.1177/0165551515613226
11. Onan, A. 2017. Hybrid supervised clustering based ensemble scheme for text classification, Kybernetes, Vol. 46, pp. 330–348. DOI: 10.1108/K-10-2016-0300
12. Onan, A. and Tocoglu, M.A. 2021. A term weighted neural language model and stacked bidirectional LSTM based framework for sarcasm identification, IEEE Access, Vol. 9, pp. 7701–7722. DOI: 10.1109/ACCESS.2021.3049734
13. Onan, A., Korukoğlu, S., and Bulut, H. 2017 A hybrid ensemble pruning approach based on consensus clustering and multi-objective evolutionary algorithm for sentiment classification, Information Processing and Management Vol. 53, pp. 814–833. DOI: 10.1016/j.ipm.2017.02.008
14. Toçoğlu, M.A., and Onan, A. 2019. Satire detection in Turkish news articles: A machine learning approach, in Communications in Computer and Information Science, Cham: Springer International Publishing, pp. 107–117. DOI: 10.1007/978-3-030-27355-2_8
15. Onan, A. 2018. Review spam detection based on psychological and linguistic features, 26th Signal Processing and Communications Applications Conference (SIU), 2-5 May, Izmir, Turkey
16. Onan, A. 2018. An ensemble scheme based on language function analysis and feature engineering for text genre classification, Journal of Information Science, Vol. 44, pp. 28–47. DOI: 10.1177/0165551516677911
17. Jahanbin, K., and Rahmanian, V. 2020. Using twitter and web news mining to predict COVID-19 outbreak, Asian Pacific Journal of Tropical Medicine, Vol. 13, pp. 378-380. DOI: 10.4103/1995-7645.279651
18. Ordun, C., Purushotham S., and Raff, E. 2020. Exploratory analysis of covid-19 tweets using topic modeling, UMAP, and DiGraphs, https://arxiv.org/abs/2005.03082 (Date of Access: 06.04.2020)
19. Peng, Z., Wang, R., Liu, L., and Wu, H. 2020. Exploring urban spatial features of COVID-19 transmission in Wuhan based on social media data, ISPRS International Journal of Geo-Information, Vol. 9, DOI: 10.3390/ijgi9060402
20. Li, D., Chaudhary, H., and Zhang, Z. 2020. Modeling spatiotemporal pattern of depressive symptoms caused by COVID-19 using social media data mining, International Journal of Environmental Research and Puplic Health, Vol. 17, DOI: 10.3390/İJERPH17144988
21. Chen, N., Zhong, Z., and Pang, J. 2021. An exploratory study of COVID-19 information on Twitter in the Greater Region, Big Data and Cognitive Computing, Vol. 5, DOI: 10.3390/bdcc5010005
22. Boon-Itt, S., and Skunkan, Y. 2020. Public perception of the COVID-19 pandemic on Twitter: Sentiment analysis and topic modeling study, JMIR Public Health and Surveill., Vol. 6, DOI: 10.2196/21978
23. Onan, A., 2021. COVID-19 ile İlgili Sosyal Medya Gönderilerinin Metin Madenciliği Yöntemlerine Dayalı Olarak Zaman-Mekansal Analizi, European Journal of Science and Technology, Vol. 26, pp. 138-143. DOI: 10.31590/ejosat.957020
24. Onan, A., and Toçoğlu, M.A. 2020. Weighted word embeddings and clustering‐based identification of question topics in MOOC discussion forum posts, Computer Applications in Engineering Education, Vol. 29, pp. 675–689. DOI: 10.1002/cae.22252
25. Onan, A. 2021. Sentiment analysis on massive open online course evaluations: A text mining and deep learning approach, Computer Applications in Engineering Education, Vol. 29, pp. 572–589. DOI: 10.1002/cae.22253
26. Bustillos, R.O., Cabada, R.Z., Estrada, M.L.B, and Perez, Y.H. 2019. Opinion mining and emotion recognition in an intelligent learning environment ,Computer Applications in Engineering Education, Vol. 27, pp. 90–101. DOI: 10.1002/cae.22059
27. Cabada, R.Z., Estrada, M. L. B., and Bustillos, R. O. 2018. Mining of Educational Opinions with Deep Learning, Journal of Universal Computer Science, Vol. 24, pp. 1604–1626.
28. Nguyen, H. T., and Nguyen, M.L. 2018. Multilingual opinion mining on YouTube – A convolutional N-gram BiLSTM word embedding, Information Processing and Management Vol. 54, pp. 451–462. DOI: 10.1016/j.ipm.2018.02.001
29. Lin, Q., Zhu, Y. , Zhang, S., Shi , P., Guo, Q., and Niu, Z. 2019. Lexical based automated teaching evaluation via students’ short reviews, Computer Applications in Engineering Education, Vol. 27, pp. 194–205. DOI: 10.1155/2021/5596518
30. López, M. B., Alor-Hernández G., Sánchez-Cervantes, J. L., Pilar Salas-Zárate M., and Paredes-Valverde, M.A. 2018. EduRP: an Educational Resources Platform based on Opinion Mining and Semantic Web, Journal of Universal Computer Science, Vol. 24, pp. 1515–1535. DOI: 10.3217/JUCS-024-11-1515
31. Chen, T., Peng, L., Jing, B., Wu, C., Yang, J., and Cong, G. 2020. The impact of the COVID-19 pandemic on user experience with online education platforms in China, Sustainability, Vol. 12, DOI: 10.3390/su12187329
32. Komasawa, N., Terasaki, F., Nakano, T., Saura, R. , and Kawata, R. 2020. A text mining analysis of perceptions of the COVID-19 pandemic among final-year medical students, Acute Medicine Surgery, Vol. 7, pp. DOI: 10.1002/ams2.576
33. Kim, E.-J., Kim J. J., and Han, S.-H. 2021. Understanding student acceptance of online learning systems in higher education: Application of social psychology theories with consideration of user innovativeness, Sustainability, Vol. 13, DOI: 10.3390/su13020896
34. Porter, M. F. 2001. A language for stemming algorithms. http://snowball.tartarus/texts/introduction (Date of Access: 05.10.2001)
35. Lane, P. C. R., Clarke D., and Hender, P. 2012. On developing robust models for favourability analysis: Model choice, feature sets and imbalanced data, Decision Support System, Vol. 53,pp.712–718. DOI: 10.1016/J.DSS.2012.05.028
36. Hackeling, G. 2017. Mastering machine learning with scikit-learn -, 2nd ed, Birmingham, England: Packt Publishing, 254p.
37. Vapnik, V. 2014. The nature of statistical learning theory. New York, NY: Springer, 314p.
38. Li, X., Li, S., Li, J., Yao, J., and Xiao, X. 2021. Detection of fake-video uploaders on social media using Naive Bayesian model with social cues, Scientific Reports, Vol. 11, DOI: 1038/s41598-021-95514-5
39. Hastie, T., Tibsharani, R., and Friedman, J. 2009. Springer Series in Statistics The Elements of, Math. Intell, Vol. 27, pp. 83–85.
40. Breiman, L., Last, M., and Rice, J. 2006. Random forests: Finding quasars, in Statistical Challenges in Astronomy, New York: Springer-Verlag, pp. 243–254.
41. Bengio, Y., and Senecal, J.S. 2008. Adaptive importance sampling to accelerate training of a neural probabilistic language model, IEEE Transactions on Neural Networks, Vol. 19, DOI: 10.1109/TNN.2007.912312
42. Rezaeinia, S. M., Rahmani, R., Ghodsi, A., and Veisi, H. 2019. Sentiment analysis based on improved pre-trained word embeddings, Expert System with Application, Vol. 117, pp. 139–147. DOI: 10.1016/j.eswa.2018.08.044
43. Mikolov, T., Chen, K., Corrado, G., and Dean, J. 2013. Efficientestimation of word representations in vector space, https://arxiv.org/abs/1301.3781.(Date of Access: 07.09.2013)
44. Joulin, A., Grave, E., Bojanowski, P., Douze, M., Jégou, H., and Mikolov, T. 2016. Fasttext. zip: Compressing text classification models. https://arxiv.org/abs/1612.03651 (Date of Access: 12.12.2016)
45. Di, W., Bhardwaj, A., and Wei, J. 2018. Deep Learning Essentials: Your hands-on guide to the fundamentals of deep learning and neural network modeling. Packt Publishing, 284p.
46. Pennington, J., Socher, R., and Manning, C. 2014. Glove: Global vectors for word representation, in Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), October 2014, Doha, Qatar, 1532-1543
47. LeCun, Y., 1989. Generalization and network design strategies, Vol. 19, Amsterdam: Elsevier
48. Elman, J.L. 2020. Finding structure in time, in Connectionist psychology: A text with readings, Psychology Press, 352p.
49. Zhang, L., Wang, S., and Liu, B. 2018. Deep learning for sentiment analysis: A survey, Wiley Interdiscip. Rev. Data Min. Knowl. Discov., Vol. 8, DOI: 10.1002/widm.1253
50. Rojas-Barahona, L.M. 2016. Deep learning for sentiment analysis: Language and Linguistics Compass, Language and Linguist. Compass, Vol. 10, pp. 701–719. DOI: 10.111/Inc3.12228
51. Cho, K., et al. 2014. Learning phrase representations using RNN encoder-decoder for statistical machine translation. https://arxiv.org/abs/1406.1078 (Date of Access: 03.06.2014)
52. Devlin, J., Chang , M.-W., Lee, K. , and Toutanova, K. 2018. BERT: Pre-training of deep bidirectional Transformers for language understanding. https://arxiv.org/abs/1810.04805 (Date of Access: 11.10.2018)