What is it about?
Recommender systems help users find relevant content almost everywhere in today's Web. Most research focuses on the algorithms, but the presentation of recommendations is at least as important from the user's perspective. For a long time, the presentation consisted of simple one-dimensional lists. In recent years, however, it has become the de-facto standard to display multiple collections of recommendations, so-called "carousels". Netflix, for example, shows different types of personalized movie recommendations based on genres, popular themes, and curated content. Spotify recommends new releases, podcasts on certain topics, and songs similar users are listening to. Despite the popularity of these multi-list recommender interfaces (MLRI), the empirical basis for their design is weak. In this paper, we attempted to fill this gap through an exploratory user study, trying to shed light on the influence of specific design aspects, such as carousel type and length, on the individual perception and usage of the interfaces.
Featured Image
Photo by heni noviyanti on Unsplash
Why is it important?
Research on carousel interfaces that actually involves users is rare. As a result, assumptions are made that have not been tested with real users, even though it is known from conventional recommendation lists that individual decision making can play a significant role in the recommendation process. Our work provides first insights into the differences in decision making in the presence of multiple recommendation carousels. Most importantly, we found that neither the composition nor the order of carousels plays a significant role. Also, the visible length of carousels seems to be of limited importance. However, the questionnaire results suggest that presenting six items per carousel – as in most real-world systems – can be considered a good compromise. On the other hand, as we observed inter-individual differences in the perception and usage of the interfaces, they cannot be considered a one-fits-all solution. Thus, it appears that there are still many unanswered questions about how to best design personalized recommender interfaces.
Perspectives
As is often the case in recommender research, the user perspective is largely ignored when it comes to the design of carousel interfaces. This was particularly surprising to me given the popularity of this presentation format. I hope that this paper can contribute to the development of a better understanding of the interaction with and usage of multi-list recommender interfaces.
Benedikt Loepp
Fraunhofer IMS
Read the Original
This page is a summary of: How Users Ride the Carousel: Exploring the Design of Multi-List Recommender Interfaces From a User Perspective, September 2023, ACM (Association for Computing Machinery),
DOI: 10.1145/3604915.3610638.
You can read the full text:
Resources
Contributors
The following have contributed to this page
