Social media mining to investigate the impacts of the COVID-19 pandemic
Loading...
Date
2022-08-16
Authors
Parsa, Mohammad S.
Journal Title
Journal ISSN
Volume Title
Publisher
University of Waterloo
Abstract
The COVID-19 pandemic created a global crisis with devastating social and economic impacts. Firstly, public health measures for COVID-19, such as social distancing affected how we work and study. Secondly, this crisis caused mobility restrictions and shutdowns that impacted our economy. In this thesis, we aim to obtain a better understanding of how these socio-economic impacts have affected people. We, therefore, choose one problem from each of these two areas of impact for further study.
The social distancing mandates shifted working environments and education online. Due to cheating being more prevalent in online education, serious issues may arise during the pandemic when classes and examinations are online. In order to understand these issues and their impacts on college students, we ask: how do college students feel about online cheating?
Fuel consumption and carbon emissions declined due to mobility restrictions and economic shutdowns. As a result, air quality improved. Economic shutdowns, however, impacted countries' ability to fight climate change. We are interested in understanding how people's perspectives have changed due to both the positive and negative effects of the pandemic on climate change. To do so, we ask: What is the public's attitude towards climate action during the COVID-19 recovery and beyond?
We answer these questions by analyzing discussions on Twitter and Reddit social media platforms. These online social media platforms are considered essential tools for reflecting and forecasting public opinion on a wide range of topics. Therefore, we answer our questions by mining text messages that were posted during the COVID-19 crisis. We begin by collecting the necessary posts and comments. We then prepare the documents for text mining by using standard pre-processing techniques. As a result, we are able to construct an understanding of the discussions by using these methods.
While investigating the discussions about academic dishonesty on Reddit, we found more discussions related to cheating in 2020 than in 2019. The topics have expanded from plagiarism in programming assignments to online assessments in general. Topic modelling of the Fall 2020 discussions revealed three concerns raised by students: that cheating inflates grades and forces instructors to increase the difficulty of assessments; that witnessing cheating go unpunished is demotivating; and that academic integrity policies are not always communicated clearly.
Investigating the discussions about the climate change and the pandemic on Twitter revealed that most tweets support climate action and point out lessons learned during the pandemic response that can shape future climate policy, although skeptics continue to have a presence. Additionally, concerns arise in the context of climate action during the pandemic, such as mitigating the risk of COVID-19 transmission on public transit.
Description
Keywords
Social Media, Text mining, Topic modeling, COVID-19, Coronavirus, Natural language processing