Research
I study how people interact with information through recommender systems and personalization. My research connects reliable experimental evidence with people’s experiences and the environmental and societal consequences of personalized systems.
Evaluation and reproducibility
What does an evaluation result allow us to conclude?
Evaluation depends on datasets, sampling, tuning and software as well as algorithms. I study how these choices shape comparisons and what evidence supports a reported improvement. This work spans reproducible benchmarking with RiVal, audits of experimental frameworks, and methods for interpreting partial or conflicting reproduction results.
Selected contributions
Reproducibility in Recommender Systems: A Survey
The Hidden Cost of Defaults in Recommender System Evaluation
Calibrating Reproduced Claims in Recommender Systems
People and personalization
What do people need from personalized systems?
I study how preferences change over time and how people understand and trust personalized systems. Through studies of music listening, explanations and conversational agents, I examine when personalization helps, how context shapes its value, and how trust develops through use.
Selected contributions
Soundtracks of Our Lives: How Age Influences Musical Preferences
User Perceptions of Personalized and Generic Explanations in LLM-Driven Recommender Systems
Trust as a Situated User State in Social LLM-Based Chatbots: A Longitudinal Study of Snapchat’s My AI
Sustainability and responsibility
What should count as a cost or benefit of personalization?
I examine the energy use and carbon footprint of recommender-system research and contribute measurement tools such as EMERS. I also study wider consequences: how personalization can encourage consumption, whose interests systems serve, and how people understand fairness criteria. These questions connect environmental measurement with system design and evaluation.