Research

I’m a PhD candidate on the topic of user control in AI interfaces. I surveyed 360 people across 46 countries and reviewed a decade of AI-UX research; the resulting papers are now under peer review at top HCI journals. The recurring finding: people want clarity, refinement tools, and better personalization - insight that shapes everything I design, including the product I lead, flero.ai.

This is the reading behind that work - every paper I’ve collected, mapped across five dimensions of AI-UX research. Browse by theme to see what each means, or switch the lens to read the papers as a grid. More about me at tanjaradovanovic.com.

AI technique22

UX effect21

Challenge17

Fragmented research50 Designing for emotion32 Adapting to diverse users28 Measuring user experience23 Bridging theory and practice21 Explainability20 Measuring emotion20 Defining emotion13 Designing for uncertainty6 Adaptation usability trade-off5 Communicating intangible/sensory qualities4 Scalability3 Stakeholder collaboration3 Disorientation2 Privacy-preserving personalization2 AI unpredictability1 Motion sickness1

Metric18

Self-report measures10 Qualitative interviews9 Survey responses8 Task performance8 Controlled experiment7 Physiological measures7 Behavioural measures5 EEG / neural measures5 Eye tracking4 Facial expression analysis4 Accuracy / precision / recall3 Presence rating3 Usability evaluation3 Emotion scales (SAM/PAD)2 A/B testing1 Engagement metrics1 Ownership & realism1 Wayfinding performance1

Ethical concern12

Accountability10 User autonomy & agency10 Privacy7 Societal impact of AI6 User empowerment6 Algorithmic bias & fairness4 Emotional manipulation4 Behaviour manipulation2 Data consent2 Filter bubbles & echo chambers2 Inclusion & accessibility2 Digital divide1