The COVID-19 infodemic : Telling Facts from Fakes

The Internet has become a major source of information, with a single piece shared across different platforms potentially reaching millions in a short period of time. As Covid-19 spreads across the world, the misinformation and fake news around it also spread. For each fact about Covid-19 made public, a large body of misinformation grows and gains traction (e.g. false origin of this disease, unproven treatments, the impact it has on different companies, governments and others (Raman Sandhya., 2020)). In response we must detect misinformation and suspicious posts around Covid-19 and prevent the potential impact on millions. In this project, we will apply machine learning and natural language processing techniques to distinguish fact from fake and classify information around Covid-19 into pre-defined categories. Furthermore, we will use intelligent web crawling strategies to iteratively gather a large diverse dataset, and ensure a robust model which can thrive in the ever changing misinformation ecosystem. Finally, we will augment our content-based model with syntactic and contextual information to even more robustly separate fact from fake.

Faculty Supervisor:

Stan Matwin

Student:

Sima Sharifirad

Partner:

Factually Health

Discipline:

Computer science

Sector:

University:

Dalhousie University

Program: