English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recommender systems (RS) have achieved significant success by leveragingexplicit identification (ID) features. However, the full potential of contentfeatures, especially the pure image pixel features, remains relativelyunexplored. The limited availability of large, diverse, and content-drivenimage recommendation datasets has hindered the use of raw images as itemrepresentations. In this regard, we present PixelRec, a massive image-centricrecommendation dataset that includes approximately 200 million user-imageinteractions, 30 million users, and 400,000 high-quality cover images. Byproviding direct access to raw image pixels, PixelRec enables recommendationmodels to learn item representation directly from them. To demonstrate itsutility, we begin by presenting the results of several classical pure ID-basedbaseline models, termed IDNet, trained on PixelRec. Then, to show theeffectiveness of the dataset's image features, we substitute the itemIDembeddings (from IDNet) with a powerful vision encoder that represents itemsusing their raw image pixels. This new model is dubbed PixelNet.Our findingsindicate that even in standard, non-cold start recommendation settings whereIDNet is recognized as highly effective, PixelNet can already perform equallywell or even better than IDNet. Moreover, PixelNet has several other notableadvantages over IDNet, such as being more effective in cold-start andcross-domain recommendation scenarios. These results underscore the importanceof visual features in PixelRec. We believe that PixelRec can serve as acritical resource and testing ground for research on recommendation models thatemphasize image pixel content. The dataset, code, and leaderboard will beavailable at https://github.com/westlake-repl/PixelRec.

An Image Dataset for Benchmarking Recommender Systems with Raw Pixels