Towards Multi-Method Support for Product Search and Recommending

Kleemann, T., Loepp, B., & Ziegler, J. (to appear). In UMAP ’22: Adjunct Proceedings of the 30th ACM Conference on User Modeling, Adaptation and Personalization. ACM.


Today, online shops offer a variety of components to support users in finding suitable items, ranging from filters and recommendations to conversational advisors and natural language chatbots. Not only do all these methods differ in terms of cognitive load and interaction effort, but also in their suitability for the particular user. However, users can often choose between a number of methods to reach their goals, making it difficult to determine which to use, and, as the settings from one component are not propagated to the others, to switch between. In this paper, we study the reasons for using the different components in more detail and present a first multi-method approach for providing a more seamless experience, where users can freely and flexibly choose from all methods at any time.


Related publications


On User Awareness in Model-Based Collaborative Filtering Systems

Interactive Recommending with Tag-Enhanced Matrix Factorization (TagMF)

More »

Focus area