Open AccessEVALUATING COVID-19 HEALTH INFORMATION USING MACHINE LEARNING
Author(s) -
Buddavarapu Teja Swaroop
Publication year - 2020
Publication title -
international journal of engineering applied science and technology
Language(s) - English
Resource type - Journals
ISSN - 2455-2143
DOI - 10.33564/ijeast.2020.v05i05.052
Subject(s) - covid-19 , computer science , artificial intelligence , virology , machine learning , medicine , disease , outbreak , infectious disease (medical specialty)
A high volume, variety and variability amount of COVID-19 information (data) is available online which might be inaccurate and incorrect. In this paper we quantify, analyse the available COVID-19 content about health guidance using ML and topic modelling, mainly concerned on vaccinations (“anti-vaccination”). It is found that the anti-vaccination community is less focused part around COVID-19 as compared to counterpart, the pro-vaccination (provaccination) community. However, the antivaccination community has a broader range of “flavours” of COVID-19 topics, so it has a broader cross-section of individual opinion and their version in COVID-19 guidance available online, e.g. some of them wary of important fasttracked COVID-19 vaccine and some developing and studying alternative remedies. Hence the anti-vaccination community looks better than the pro-vaccination community as has more support and awareness in the society. This is increasing since a widespread lack of awareness and adoption of a vaccine which makes the world to face in fall rapidly in providing herd immunity, which results in leaving countries open to future resurgences. We developed a ML model that analyse these data and interprets results and helps in assessing the efficiently provide strategies. In this paper the approach done is a scalable and so, tackles the problem facing due to social media platforms which is spreading variety, various and huge amount of misinformation and disinformation which may panic public and this paper is not resembling that all information available are not of these type but most of them are. KeywordsMachine learning, topic modelling, social computing, medical information systems, community.