CAREER: Enhanced Analysis & Algorithms to Minimize the Spread of Misinformation in Social Networks

The project goal is to significantly reduce the destructive spread of misinformation on social media and other Web sources, and its potential threat to national security. To do so, the investigator will develop new machine learning algorithms to better detect authenticity and recommend content. The research integrates computer and social sciences to account for the complex real-world interactions among publisher, platform, content recommendation algorithms, bot users, human users, and their social connections. One way social networks engage users is by keeping them consuming personalized content. Malicious actors can easily penetrate these systems with misleading stories and consequent recommendations that prompt people to make decisions based on this misinformation. Younger generations are increasingly active on such platforms, making it critical to reduce the threat that the continuing spread of misinformation poses. Findings will result in a deeper understanding of how recommender systems behave in the presence of misleading stories, and will offer systems design strategies to insure that people receive accurate information to make decisions.

There is currently no framework in place to quantify how much recommendations with misinformation in the loop influence social network users. This research will fill this gap. Project objectives are to: (1) develop graph-based models to measure the degree of story, sources, and user credibility as opposed to a typical binary assessment; (2) develop a new framework integrating user-centric information diffusion models to assess the impact of, and compute benchmarks for, current recommender systems in spreading misleading stories; (3) develop algorithms for content recommender systems that will minimize misinformation spread in social networks. An integrated education plan will engage Boise State University college students, who will use a service-learning approach to help Idaho high school students and teachers improve their ability to identify and respond to misinformation. Educational activities will also increase awareness of and interest in computer science occupations, and encourage minorities' retention and diversity.

This project is jointly funded by Secure and Trustworthy Cyberspace (SaTC) program and the Established Program to Stimulate Competitive Research (EPSCoR).

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.