A Proposed Sentiment Analysis Deep Learning Algorithm for Analyzing COVID-19 Tweets

Abstract

With the rise in cases of COVID-19, a bizarre situation of pressure was mounted on each country to make arrangements to control the population and utilize the available resources appropriately. The swiftly rising of positive cases globally created panic, anxiety and depression among people. The effect of this deadly disease was found to be directly proportional to the physical and mental health of the population. As of 28 October 2020, more than 40 million people are tested positive and more than 1 million deaths have been recorded. The most dominant tool that disturbed human life during this time is social media. The tweets regarding COVID-19, whether it was a number of positive cases or deaths, induced a wave of fear and anxiety among people living in different parts of the world. Nobody can deny the truth that social media is everywhere and everybody is connected with it directly or indirectly. This offers an opportunity for researchers and data scientists to access the data for academic and research use. The social media data contains many data that relate to real-life events like COVID-19. In this paper, an analysis of Twitter data has been done through the R programming language. We have collected the Twitter data based on hashtag keywords, including COVID-19, coronavirus, deaths, new case, recovered. In this study, we have designed an algorithm called Hybrid Heterogeneous Support Vector Machine (H-SVM) and performed the sentiment classification and classified them positive, negative and neutral sentiment scores. We have also compared the performance of the proposed algorithm on certain parameters like precision, recall, F1 score and accuracy with Recurrent Neural Network (RNN) and Support Vector Machine (SVM).

Publication DOI: https://doi.org/10.1007/s10796-021-10135-7
Divisions: College of Business and Social Sciences > Aston Business School
College of Business and Social Sciences > Aston Business School > Operations & Information Management
Funding Information: This research work is catalyzed and supported by the National Council for Science and Technology Communication (NCSTC), Department of Science and Technology, Ministry of Science and Technology (Govt. of India), New Delhi, India (Grant Recipient: Dr. Harle
Additional Information: © Springer Nature B.V. 2021. The final publication is available at Springer via http://dx.doi.org/10.1007/s10796-021-10135-7 Funding Information: This research work is catalyzed and supported by the National Council for Science and Technology Communication (NCSTC), Department of Science and Technology, Ministry of Science and Technology (Govt. of India), New Delhi, India (Grant Recipient: Dr. Harleen Kaur). This work is partly supported by VC Research (VCR 000116) for Prof. Chang.
Uncontrolled Keywords: COVID-19,Heterogeneous Euclidean overlap metric (H-EOM),Hybrid heterogeneous support vector machine (H-SVM),Recurrent neural network (RCN),Sentiment analysis,Twitter,Theoretical Computer Science,Software,Information Systems,Computer Networks and Communications
Publication ISSN: 1572-9419
Last Modified: 27 May 2024 07:38
Date Deposited: 07 Jun 2022 15:02
Full Text Link:
Related URLs: http://www.scop ... tnerID=8YFLogxK (Scopus URL)
https://link.sp ... 796-021-10135-7 (Publisher URL)
PURE Output Type: Article
Published Date: 2021-12-01
Published Online Date: 2021-04-20
Accepted Date: 2021-04-06
Authors: Kaur, Harleen
Ahsaan, Shafqat Ul
Alankar, Bhavya
Chang, Victor (ORCID Profile 0000-0002-8012-5852)

Download

[img]

Version: Accepted Version

| Preview

Export / Share Citation


Statistics

Additional statistics for this record