The past decade has seen the meteoric rise of social media – and with it, fake news.
Encompassing everything from health to politics, online disinformation can spread rapidly, and is becoming harder to detect.
Enter University of Southern Queensland computer scientist Dr Taotao Cai, who is on a mission to sort the real from the fake.
Dr Cai is preparing to launch a new project which will centre on creating an algorithm to detect fake news and track its spread on social media.
This follows on from his previous work studying Influence Maximisation algorithms, a system which identifies individuals on social media that hold the most influence.
“We carried out research on user engagement in communities, with an emphasis on pinpointing the influential users who play a crucial role in shaping social communities,” Dr Cai said.
“To achieve this, we developed a suite of sophisticated algorithms to identify key users who can inspire fellow members to maintain active participation in the community, ultimately leading to its expansion.
“This refined strategy enhances information dissemination within the emerging community.”
Dr Cai said he intends to incorporate this knowledge of information diffusion to help better predict the spread of fake news.
“Fake news poses a threat to society and there are many real-world examples of it being harmful,” Dr Cai said.
“Previous studies into fake news have focused mainly on methods of truth detection, which can identify real information from the false - we are now looking to combine this with a diffusion model.
“Often these models have a gap between predictions and real-world outcomes.
“To combat this, I am looking to merge traditional diffusion models with current graph neural network (GNN) technology (a machine learning algorithm that tracks connections).”
His findings, combined with others in the field, will help prevent the spread of incorrect information online.
“Our foremost aim is to construct precise models for tracking the diffusion of false information, while also creating cutting-edge technologies and algorithms to combat fake news,” he said.
“This approach provides a robust foundation for effectively mitigating the spread of misinformation.
“By employing sophisticated key path search algorithms to detect and obstruct the dissemination channels of deceptive content, we can substantially reduce its prevalence in social networks, ultimately protecting the public from the negative impacts of misinformation."
Learn more about the University’s School of Mathematics, Physics and Computing.