ARLINGTON, VA, October 5, 2020 – Quantitative Scientific Solutions today announced that it has been selected by the Office of Naval Research’s Small Business Innovation Research (SBIR) program to develop a suite of advanced machine learning algorithms and an associated desktop extension and mobile phone app to help identify large classes of bots, bot-assisted accounts, human propagandists, purveyors of disinformation, and other malicious accounts. This Sociolinguistic Information Filtering Tool (SIFT) project will enable social media users to “fight back against the algorithms” and block not only manipulation and disinformation, but also to gain fine-grained control over the types of accounts and content they would like in their feed.
Current solutions for addressing malicious accounts are inadequate, focusing on bots and basing analysis on variables derived from simple text features and account metadata. SIFT will significantly innovate on these approaches by leveraging novel methods from computational linguistics to analyze patterns between accounts and performing a holistic mathematical analysis of social network structures to determine malicious actors with greater accuracy.
QS-2 has teamed with The Pennsylvania State University to augment SIFT’s capabilities using novel game theoretic modeling and user testing of the tool based on social science.
QS-2’s Director of Analytics and Founding Partner, Dr. David Guarrera said,
“With the current proliferation of misinformation, disinformation, bots, deepfakes, and other manifestations of purposefully manipulative content, new tools are needed now to regain trust in the online public discourse and combat coordinated online manipulation at scale. SIFT’s innovations in linguistic methods, large social network analysis, and game theory will be crucial in developing an effective app for robust and early detection of malicious accounts. SIFT will add an essential weapon in fighting increasingly worrying online trends.”