Description

The COVID-19 pandemic has escalated to an unprecedented global concern and was declared a Public Health Emergency on January 30, 2020, by the World Health Organization. The crisis has already threatened almost every country's financial stability and had a devastating effect on the global economy.

Considering the unmanageable situation with the spread of the virus, most countries have adopted a lockdown policy, in which travel restrictions are applied, most business activities are forced to close down, and people are asked to stay at home.

In this context, it is essential to evaluate and investigate how societies react to the pandemic's problematic situation in different countries in Western Europe. Additionally, it is crucial to know how people respond to government's suggestions, recommendations, and rulings that usually restrict their everyday activities.

Social media can be considered a place for measuring different communities' pulse. People share insights, feelings, opinions, and worries while disseminating health events' interpretations on social media (Khan, Fleischauer, Casani, & Groseclose, 2010). The changes in the language that people use during a crisis can provide interesting information about their opinions. Moreover, the use of mobile devices equipped with positioning sensors enriches the messages with spatio-temporal information. This unstructured, spatio-temporal information collected from people is a suitable source to analyze the mob's reaction to the pandemic in different locations, which is essential in outbreak surveillance efforts.

This study's primary goal is to contribute to the current COVID-19 surveillance in Western Europe by monitoring and analyzing the main topics of concern about COVID-19 over time and space, using Twitter as the most popular microblogging social network. Machine Learning and Natural Language Processing (NLP) techniques have been used previously by other researchers to extract spatio-temporal information from Twitter feed (see Farnaghi, Ghaemi, and Mansourian (2020) and Ghaemi and Farnaghi (2019)). These techniques will be used to extract essential conversations about COVID-19, temporally, throughout the course of the pandemic, and then spatiotemporally to highlight the regions where people were more apprehensive about the pandemic. You will link the finding to the frequencies of COVID-19 cases in different countries and discuss the results.Dataset

A huge dataset of geotagged COVID-19 tweets collected between February 1 and December 31, 2020 is available for this study.Required skills

Good knowledge of Python programming language will be helpful for this MSc topic.

Objectives and Methodology

To monitor and analyze the dialog about COVID-19 on Twitter over time and space in Western Europe using Geotagged tweets and investigate the relationship between people's attention to the pandemic and the disease's spread.

Further reading

Farnaghi, M., Ghaemi, Z., & Mansourian, A. (2020). Dynamic Spatio-Temporal Tweet Mining for Event Detection: A Case Study of Hurricane Florence. International Journal of Disaster Risk Science, 11(3), 378-393. doi:10.1007/s13753-020-00280-z
Ghaemi, Z., & Farnaghi, M. (2019). A Varied Density-based Clustering Approach for Event Detection from Heterogeneous Twitter Data. ISPRS International Journal of Geo-Information, 8(2), 82.
Khan, A. S., Fleischauer, A., Casani, J., & Groseclose, S. L. (2010). The next public health revolution: public health information fusion and social networks. American journal of public health, 100(7), 1237-1242. doi:10.2105/AJPH.2009.180489