Virginia Tech® home

An Assessment Framework for Information-Handling Practices on Social Media

Researchers will design, develop, and evaluate a living, learning artificial intelligence-based framework through development of intelligent data-centric assessment tools to identify practices that could lead to disinformation’s spread on social media.

Project funded by: CCI Hub

Project Investigators

Principal Investigator (PI): Tamer Nadeem, associate professor, Virginia Commonwealth University Department of Computer Science

Co-PI: Jeanine Guidry, assistant professor, Virginia Commonwealth University School of Media and Culture

Rationale and Background

The viral spread of digital misinformation, especially through social media platforms, has become so serious that the World Economic Forum considers it a threat to human society. 

Training and educating people on how to recognize misinformation, verify sources, and make responsible decisions when sharing information is of paramount importance.

Methodology

The target population will be students, both in high school and college, as they’re one of the earliest groups age-wise to intensively use social media and potentially share misinformation. 

The project scope will focus on health-related misinformation on the Twitter platform by:

  • Identifying information-handling engagement practices of individual users that may lead to spreading disinformation.
  • Identifying the class of practices and users that may lead to spreading misinformation.

The assessment will be done through monitoring subjects’ interactions and responses to social media activities with tools that are integrated seamlessly into their daily social media activities so as not to interfere with typical interactions and affect typical behavior. Steps include:

  • Conducting a survey to understand stakeholders’ activities/actions, concerns, needs, etc. with respect to teaching students how to properly respond to and make decisions about sharing online health information. 
  • Hosting six virtual focus groups among stakeholders (including school and higher education officials and instructors, academics, social media platform leaders, and government/public health officials) to get in-depth clarification on the tool characteristics.
  • Developing intelligent data-centric assessment tools to identify bad practices of individuals on social media that would lead to information dissemination through monitoring an individual’s interactions and responses during activities with information on social media.
  • Determining if there are differences in demographics, perceptions, and social media engagement between individuals who share misinformation messages and those who first verify the information and only share the non-misinformation messages.

Projected Outcomes

  • Development of a deep-learning classification tool to identify the signature/profile of individuals, which could be used by other organizations to identify employees who need training on handling information.
  • Strategies to train people to evaluate, recognize, and deal with health-related misinformation before they find it on social media.
  • Research results will be disseminated via journal/conference submissions and workshop attendance.
  • Some results will be further developed as training and course materials, which will instantly benefit workforce development. 
  • Developed tools and collected data will be made available to the CCI network for training, teaching, and research purposes.