The Vibrent Clothes Rental dataset was created by researchers at the University of Agder in 2024. The related paper result is "A Dataset for Adapting Recommender Systems to the Fashion Rental Economy", which has been accepted by RecSys 2024.

The dataset comes from the research team's collaboration with a Norwegian fashion rental company and contains 64k transactions, rental histories of 2.2k anonymous users, and 15.8k unique outfits, with detailed information about the properties and rental history of each item. All clothing is listed as a single item or its corresponding item group, referring to the shared design between single items, and each item is accompanied by a set of labels describing some of its properties. The research team also provided a total of 50k images showing all items and a set of pre-computed zero-shot embeddings.

The team applied a number of common recommender system methods to the dataset to provide a performance benchmark. This benchmark was performed on both the traditional fashion recommender system problem of recommending groups of clothing and the new problem of predicting rentals for individual items. This is the first publicly available dataset discussing the clothing rental domain, and the research team hopes that the release of this dataset will serve as a catalyst for a new branch of research dedicated to fashion rental recommender systems.