Predicting what they will share

Blog post number 4 in our 12 days of HPC series from School of Languages!

During the month of December we’re featuring blog posts from researchers from across the University of Leeds showcasing the fantastic work they do using our High Performance Computing system. Follow us @RC_at_Leeds to keep up to date with our 12 days of HPC blog series.

What’s your name?

Serge Sharoff

What department do you work in?

School of Languages

What research question are you trying to answer?

The project aims to improve our understanding of how COVID-19 misinformation leads to confusion, anti-health policy sentiment and risk-tolerant behaviour by examining misinformation chains in mainstream and social media using AI tracing tools. Our main hypothesis is that the socio-demographic profile of the audience is an important indicator for the likelihood of falling prey to misinformation, as different readers differ in how they might be willing to share misinformation of different kinds. For example, while there is no clear statistical difference between the older and younger generations in the number of misinformation pieces they share, there is a difference between them in their preference to share fake news vs fake treatment advice.

What tools or technologies do you use in your research? (Programming languages, packages, APIs)

Python, pytorch, Transformers

How does HPC help your research?

We collect and analyse millions of social media posts spreading COVID-19 misinformation in Twitter and Telegram. On that scale, our data collection and statistical analysis pipelines cannot work without the use of the HPC facilities, including large storage space and GPUs.

What is the potential impact of your research?

Social media is playing a significant role in disseminating public health messages with the potential to influence mistrust, misinformation and cause low confidence in public health messaging. We want to achieve better understanding for conditions of message distortion to detect and break the impact of the online misinformation chains.

In your personal opinion what’s the coolest thing about your research?

This study uses AI methods in an ethical way to uncover interesting and unpredicted patterns in Big Data coming from social media. It links properties of the message with properties of the user profiles to improve our understanding of information flows in social media with the aim of helping the public health services in responding to misinformation challenges in a more targeted way.

In your opinion, what is the ultimate Christmas song?

I Believe in Father Christmas by Greg Lake and Peter Sinfield