David Bridges
Twitter’s looking to expand its partnerships with external academic research teams in order to identify large-scale patterns in platform manipulation, with the company today opening up applications to become part of its Twitter Moderation Research Consortium (TMRC), a selected group of researchers that are regularly given access to tweet datasets related to state-backed information operations around the world.
The TMRC was launched back in June, with Twitter approving a collection of academic research teams ‘with proven records of research on content moderation and integrity topics’ who were given access to large-scale tweet datasets related to identified state-based activity. The researchers then work to examine and identify trends that can inform future approaches in combating this type of manipulation.
As explained by Twitter:
“By providing academics and researchers with access to specific, granular data (not just aggregated reports), we enable them to find insights and contextualize information in a way that increases the visibility of the reports themselves.”
Since June, members of the TMRC have shared various research reports on key Twitter usage trends based on these datasets, which have provided Twitter, along with other platforms, with key insight to better inform their enforcement efforts.
“This data has already enabled critical, independent research by TMRC’s partners at the Stanford Internet Observatory, the Cazadores de Fake News, and the Australian Strategic Policy Institute.”
In order to maximize these benefits, Twitter will now open up applications to global researchers ‘from across academia, civil society, NGOs, and journalism’.
“Our goal is simple: prioritize transparency by sharing more data on more issues to those who are studying content moderation.”
The program will ideally help to pinpoint more trends and habitual behaviors within this type of activity to advance Twitter’s enforcement efforts, which, again, Twitter can also share with other platforms as they work to crack down on manipulation.
Which is crucially important – as we’ve seen in the US, India, and more recently in regards to Ukraine, among various other regional conflicts, social media misinformation has become a key weapon in influencing public opinion, with Twitter playing a significant role in each.
Twitter’s particularly important in this respect because of its popularity among journalists – so while Twitter may have a smaller total user count than, say, Facebook, which provides broader, direct reach to the general public, Twitter is often an opinion leader, with tweeted claims getting re-amplified by key influencers and voices.
Examining these datasets in retrospect obviously won’t help to identify key issues in real time, but it may provide valuable notes on what to look for in future, and expanding that capacity to more research teams will help to maximize this effort.
Of course, providing more of this insight to more researchers can also leave that data susceptible to misuse, ala Cambrige Analytica. But Twitter also notes that approved TMRC members must have approved plans and systems in place for safeguarding the privacy and security of the data, as outlined by TMRC. Consortium members are also required to sign a data use agreement.
You can learn more about Twitter’s Moderation Research Consortium here.
Troy Sanders
Maya Kensington
