Analysis of COVID-19 pandemic using Twitter activity

Semester project, HES-SO // University of Applied Sciences Western Switzerland, 2021

  • 1 student to supervise.
  • The student worked 4 hours a week for 3 months.
  • Approximately 1 meeting per month.

Context of the project

The pandemic of COVID-19 hit the world in 2020. In collaboration wit, the Federal Office of Public Health in Switzerland, The HumanTech Institute built a platform abe to collect the activity on Twitter based on selected keywords. These keywords were chosen to get a maximum of tweets related with Covid. A machine learning classifier is implemented and can automatically label if a tweet is rather relevant (e.g. the person had the virus or symptoms) or not. In this context, the work of the student was to add the possibility to choose the machine learning algorithm used to train the model. The view and the back-end of the platform have to be modified and an additional algorithm needs to be implemented.

Tasks done by the students

  • Analysis of the existing architecture
  • Selection of files that needs to be modified
  • Mock-up of the front-end for the choice of algorithm before training
  • Implementation of this new functionality
  • Functional tests