Research
My current work sits at the intersection of trustworthy media, security & privacy, and human-centered computing — asking how manipulated content spreads and is perceived, how systems can be made more trustworthy, and how technology can be designed with communities in mind rather than around them.
Explored how deepfake-enabled profiles shape user perception of disinformation, and how AI-generated videos can be better detected and understood — including a training strategy that improved human detection of manipulated images.
Studied training strategies that enhance people's ability to spot deepfake images, and the human factors behind adoption of protective behaviors like two-factor authentication — work that produced a 60% boost in 2FA adoption and a 20% reduction in susceptibility to online disinformation.
Security analysis of password vault applications, and a case study of runtime integrity attestation (Keylime) for cloud and edge environments — bridging systems security with real-world deployment challenges.
Investigated whether explaining why a fake photo is fake actually helps people — testing explanation strategies for AI-driven content moderation and detection tools.
Designed a culturally relevant humanoid robotic dialogue agent (CultureBot) and character development strategies for storytelling robots, and studied how connected-spaces dashboards foster maker identity in afterschool makerspaces.
Applied classification machine learning to detect and categorize faults in Industrial Control Systems, improving the reliability of power-grid protections while balancing technical and user-centered needs.