Fake news, misinformation and disinformation are being created and circulated online with unprecedented speed and scale. There are concerns that this poses a serious threat to our modern digital societies by skewing public opinion about important issues and maliciously interfering with national election campaigns.
Fact-checking is an increasingly vital approach for tackling the rapid spread of false claims online. Specifically, there is an urgent need for automated systems that detect, extract and classify incorrect information in real-time; and linguistic analyses of argument structure, entailment, stance marking, and evidentiality can assist the development of such systems.
We want to bring together people with different kinds of expertise to develop new approaches for tackling the problems posed by fake news, misinformation and disinformation. Taking an existing automated fact-checking system as a baseline, the main hackathon task will be to find ways of improving its performance. The experimental framework will be used for the FEVER: Fact Extraction and VERification challenge.
Is it for me?
The task of dealing with false claims online is necessarily an interdisciplinary task. Therefore, this hackathon will create a collaborative environment for participants from a variety of backgrounds to come together to work in teams. Whether you already have strong coding skills, a specific interest in disciplines such as information engineering or natural language processing, familiarity with linguistic theory, or even an interest in the philosophy of language, you will certainly be able to make valuable contributions during the hackathon!
In particular, we encourage undergraduates and postgraduate students:
- in Engineering / Computer Science, with good programming skills (esp. Python)
- in Linguistics / Philosophy / Psychology / Sociology
- with an interest in language-based AI technologies
Do I need to be able to code?
There will be a variety of ways to get involved and contribute during the hackathon, so coding experience is not essential. For instance, participants with a background in linguistics can analyse the linguistic data in detail, and then work together with coders so that their insights can improve the baseline system.
For those participants who would like to learn more about coding, there will be introductory sessions on Python during the hackathon – so this will be a good opportunity to dip your toe in the water!
Why should I attend?
- A chance to collaborate in interdisciplinary teams to address a language-based technology problem that has huge contemporary importance.
- An opportunity to learn about the challenges of developing an automated fact-checking system, and benefit from advice and insights from fact-checking experts.
- A chance to learn Python, if you are new to coding.
The event runs from Friday to Sunday and attendees are expected to participate throughout. Lunch will be provided on all three days, and there will be coffee and snacks throughout the hackathon, to keep you going! If you have any questions about the event or would like to discuss any specific requirements please contact Shauna Concannon.