PandemicGR
Description
WHO announced the COVID-19 outbreak as a global health emergency, and later as a pandemic. The outbreak of the coronavirus leads to an outbreak of pandemic information as well in major online social media platforms, such as Twitter, Instagram and Facebook. Such pandemic information diffusion over online social media can strongly influence people’s behaviour, and thus have an impact on the effectiveness of control and protective measures deployed by the governments.
The PandemicGR project will address of the challenge of understanding and analysing the information diffusion mechanism in online social media during the COVID-19 pandemic, based on a newly collected and properly anonymised Twitter dataset concentrating on Luxembourg and the greater region.
Our analysis of the COVID-19 pandemic information on Twitter will have an immediate impact on understanding
how our society in the greater region reacts to such information during the period of public crisis and achieving
more insights on what (for both reliable information and misinformation) are their underlying dynamics and
spreading mechanisms. In the middle term, continuously monitoring and analysing the COVID-19 pandemic
information in Twitter will help to assess the consequences when the protective measures are gradually lifted and
contribute to evaluate the impact of the COVID-19 pandemic on the society and economy of the greater region.
Objectives
In this project, we aim to perform in-depth analysis of pandemic information dynamics, based on a dataset collected from Twitter users with locations labelled in Luxembourg and the greater region. In particular, we are interested in analysing user engagement and communication patterns during this public health crisis; and finding how the COVID-19 cases in these geographic areas are correlated with the Twitter conversations via a spatio-temporal analysis of the collected tweets. Since both reliable information and misinformation are propagated in online social media, thus it is of great importance to improve the trustworthiness of information related to the COVID-19 pandemic. We will develop a machine learning model to understand, model and predict COVID-19 information cascades over the social network, and a classification model to detect misinformation with high precision and recall.
Team Members
The following members are involved in the project:
- Dr. Jun Pang (Principal Investigator)
- Dr. Tegawende Bissyande (Co-Principal Investigator)
- Ninghan Chen
- Zhiqiang Zhong
Financial Support
The project is supported by the University of Luxembourg and the National Research Fund Luxembourg (grant no. COVID-19/2020-1/14700602).
