Shared Embedding Pools Let Recommenders Capture Users’ Many Facets
A new recommender framework replaces single user and item embeddings with shared multi-aspect pools, improving accuracy across six datasets.
A new recommender framework replaces single user and item embeddings with shared multi-aspect pools, improving accuracy across six datasets.
A new semantic-guided graph attention framework improves rating prediction and ranking on Amazon cross-domain recommendation pairs while showing controlled but ...
A study across five Amazon Reviews 2023 categories found that semantic popularity control reduced exposure to broadly popular content more ...
Researchers at the Vellore Institute of Technology have built NEXUS-Rec, a five-signal neural recommender that outperforms strong baselines on MovieLens ...
Researchers have developed SAGMR, a graph-learning recommender that explicitly models artistic style to improve digital media art suggestions, with the ...
A comprehensive review in Artificial Intelligence Review argues that reinforcement learning can overcome the static limitations of traditional recommender systems, ...
A new study shows that fusing browsing, clicking, carting and purchasing signals into a shared temporal representation space improves recommendation ...
© 2026 Scienmag - Science Magazine
© 2026 Scienmag - Science Magazine