The UCERSTI workshop tries to bridge the gaps between recommender systems, human computer interaction and marketing/decision-making research by providing a platform for Human-Recommender Interaction research.
In his keynote speech at the 2009 RecSys conference, Francisco Martin indicated that the main challenge in recommender system industry is not to discover algorithms that provide good recommendations, but to provide users with a usable and intuitive interface for presenting these recommendations and eliciting feedback.
Unfortunately, the research on ‘Human-Recommender Interaction’ is scarce. While algorithm optimization and off-line testing using measures like RMSE are standard procedure in the RecSys community, theorizing about consumer decision processes and measuring user satisfaction in online tests is much less common.
Meanwhile, researchers in Marketing and Decision-Making have been investigating consumer choice processes in great detail, but only sparingly put this knowledge to use in technological applications. Likewise, the field of Human-Computer Interaction has been studying the usability of interfaces for ages, but does not seem to connect the dots between research on consumer choice, and recommender system interfaces.
Read our workshop summary for more details.
Our full-day workshop includes seven paper presentations, two invited talk and a plenary discussion. Everyone is welcome to attend this workshop.
The papers will be presented at this website soon after the camera-ready deadline, and will be published using the CEUR-WS electronic publication system. Accepted workshop papers cover the following topics:
- Design and evaluation of recommender system user interfaces
- Preference elicitation methods
- User-adaptive recommender interfaces
- Qualitative evaluation of recommender systems such as case studies and think-aloud evaluations
- Quantitative evaluation of recommender systems such as controlled experiments and field trials
- User-recommender interaction measurement techniques such as questionnaires and process data analysis
- Design guidelines for recommender systems