In our project, Co·Insights, we seek to empower local communities to identify and mitigate the most pernicious misinformation online. In line with Meedan’s vision statement, we seek a world in which all people regardless of their languages, locations, and incomes have the ability to effectively locate the most relevant and pertinent information, evaluate the quality of that information, and make the decisions they want.
Funded with $5 million from the National Science Foundation’s Convergence Accelerator, we aim to make that vision a reality. In close partnership with Asian-American and Pacific Islander (AAPI) communities in the United States, we are building the knowledge, collaborations, and software needed to empower community leaders to effectively identify and counter misinformation online.
Our consortium brings together community leaders, academics, and technologists from Meedan, a technology nonprofit building software and programmatic initiatives to strengthen journalism and media literacy; the Annenberg Public Policy Center of the University of Pennsylvania and its FactCheck.org project; the University of Massachusetts Amherst, University of Connecticut, University of Colorado Boulder, and Rutgers University; AuCoDe, an AI start-up; DesiFacts; Piyaoba; Tayo; and Viet Fact Check.
This project follows a year of in-depth prototyping and user interviews in Phase I of the Convergence Accelerator. The project leverages in-depth ethnography, discourse analysis, and communication design alongside public policy, computer science, and computational social science.
Our easy-to-use, mobile-friendly tools will allow AAPI community members to forward potentially harmful content to tiplines and discover relevant context explainers, fact-checks, media literacy materials, and other misinformation interventions. Our Phase I research shows that misinformation claims often form around common themes, persistent stereotypes, and patterns of deception. By building taxonomies and using machine learning to map claims to them, we can move to a proactive model where interventions are available before claims spread widely. The Annenberg Public Policy Center has pioneered and validated this approach with infectious disease, and our team has built machine learning approaches to accurately map claims to this taxonomy. In addition to community tiplines, we will crawl and analyze AAPI-specific content using machine learning to detect controversy and match similar claims. The insights from this data can enable community leaders and fact-checkers to create more effective, targeted misinformation interventions tailored to the needs of their communities.
Please subscribe to Meedan’s newsletter, The Checklist, for project updates.
- Online conversations are heavily influenced by news coverage, like the 2022 Supreme Court decision on abortion. The relationship is less clear between big breaking news and specific increases in online misinformation.
- The tweets analyzed were a random sample qualitatively coded as “misinformation” or “not misinformation” by two qualitative coders trained in public health and internet studies.
- This method used Twitter’s historical search API
- The peak was a significant outlier compared to days before it using Grubbs' test for outliers for Chemical Abortion (p<0.2 for the decision; p<0.003 for the leak) and Herbal Abortion (p<0.001 for the decision and leak).
- All our searches were case insensitive and could match substrings; so, “revers” matches “reverse”, “reversal”, etc.