Xiao Liu

Stopping the spread of ‘fake news’ is good news in any language

• Creating a new system to detect fake news in multiple languages
• Key tool for governments as well as mainstream and social media organisations
• Potential solution to the ‘infodemic’ facing Culturally and Linguistically Diverse communities

Learn more about the other researchers who have also been awarded a Study Melbourne Research Partnerships program grant in 2021.

Since the beginning of the global pandemic, a massive amount of COVID-19 information has spread around the world through mass media and social platforms, creating what authorities term an ‘infodemic’. Unfortunately, not all information is accurate or trustworthy. Some is even categorised as misinformation or fake news.

Each country has different strategies to control COVID-19, which leads to a huge amount of inappropriate news sharing. These include links to technologies like 5G mobile networks, and unconfirmed treatments and preventative measures like using onions or the sun for protection. Meanwhile, fake news has been produced in various languages allowing messages to spread more easily in particular among ethnic groups.

Clearly, the detection of multilingual COVID-19 fake news is essential for countries with hundreds of ethnic and linguistic groups such as Australia and Indonesia. Unfortunately, existing fake news detection methods remain too general to be effective in multiple languages.

This collaboration between Victoria and Indonesia will use a computational-based detection, machine learning, and human factor engineering approach to design a decision support system that functions as a warning mechanism for misinformation and fake news related to COVID-19. The final tool is being developed for use by organisations and governments in developing appropriate communications and education strategies.