The ubiquity of social media has opened the floodgates to fake news and its immediate and enduring consequences on our democracy, economy, health, and environment as most people are relying on social media as their predominant news source. The detection of fake news remains challenging due to its dynamically changing nature, the volume of the content, and its incredibly fast-paced diffusion on social media. Fake news has been examined by applying machine learning and/or AI together with natural language processing techniques. However, previous systems typically only give a final decision as to whether the news under investigation is fake or not, with little explanatory information revealed about why the decision is made. Such a coarse dichotomy does not account for the nature of fake news being a combination of mis/disinformation mixed with factual information, and ignores rich and highly contextualised psycho-social and cognitive factors. In this project, we will develop an explanatory model for fake news detection through analysing news items from different perspectives, including attribute information (news context, speaker, etc.), news semantic meaning, and linguistic features of news statements.
You are expected to completed COMP9444 and MATH2089 (desirable) to conduct the research in this project.
The research team for this project consists of Dr Jiaojiao Jiang from CSE and an undergraduate student from UNSW. Dr Jiaojiao is an experienced researcher in the area of Cybersecurity and Deep Learning. She has published research articles on top-tier venues, including TIFS, TPDS, TDSC, etc. She has two GPU workstations for research students to conduct experiments and collect real-world datasets.
The expected outcomes from this project is a demo system that can assists end-users to identify news credibility.