Algorithmic Affordances
In this project, we delve into the user experience of recommender systems and the ways in which algorithmic affordances, or interface mechanisms that allow users tangible control over algorithms, impact interaction qualities such as transparency, trust, and serendipity. Algorithmic affordances in this context include, for example, the ability to change which data are used or the order of recommendations. Because the precise nature of the relationship between algorithmic affordances, their use in recommender interfaces, interaction qualities, and user experience is unclear, our goal is to systematically analyse related empirical research in the field of human-computer interaction and, as a result, define proper guidelines that designers and researchers can follow when working on algorithm-driven projects. These guidelines will take the form of a rich example library illustrating design considerations and academic research on various recommender interface designs, bridging the research-practice gap.
Researcher
Aletta Smits, Researcher
Lectoraat Human Experience & Media Design (HEMD)
Academic Supervisors
Koen van Turnhout (HU)
Other Partners
The Hague University of Applied Sciences, Greenberry, Blue Field Agency, MoreApp, Zeta Alpha
Publications
- Hekman, E., Nguyen, D., & Stalenhoef, M. (2022). Towards a Pattern Library for Algorithmic Affordances. In CEUR Workshop Proceedings (Vol. 3124, pp. 24-33). CEUR WS.
- Smits, A., Bartels, E., Detweiler, C., & Van Turnhout, K. (2023). Algorithmic Affordances in Recommender Interfaces [Conference workshop]. Interact 2023, York, England. https://hemdmissies.nl/interact2023/