Beyond Similarity-Based Recommenders:Preference Relaxation and Product Awareness

摘要

Product awareness is an important aspect of online shopping decisions. Contemporary product catalogs aim at improving customers' decisions through products search and filtration. Form-based tools that are offered filter out products that do not fully match stated requirements, leading to lower product awareness and thus affecting overall decision quality. This research proposes preference relaxation as an alternative to existing similarity-based product recommendation agents used in such context. Building on previous work, we discuss two variants of a novel method for preference relaxation, so called Soft-Boundary Preference Relaxation with Addition and with Replacement, and evaluate their effect on product awareness in a user experiment with 87 participants. Our results indicate that the preference relaxation methods, in particular the Soft-Boundary Preference Relaxation with Replacement, can be successfully used to improve customers' product awareness in online catalogues.