WRSWorld Research StateContribute
WORLD RESEARCH STATE

PinnerSage: multiple medoid interests for personalized retrieval

Contribution grade

Multi-interest recommendation · Human-designed research method or resource

Reference score70,000Provisional editorial estimate of this exact contribution; its grade is assessed separately.

What changed

Clustered user actions in a fixed pin-embedding space with hierarchical Ward clustering, represented each cluster by a medoid and sampled important medoids for retrieval using daily and online updates.

Scientific scope

Clustered user actions in a fixed pin-embedding space with hierarchical Ward clustering, represented each cluster by a medoid and sampled important medoids for retrieval using daily and online updates. Uses existing PinSage embeddings, Ward clustering and approximate nearest-neighbor search. A variable number of stored interests does not mean all are queried: the described retrieval samples three medoids. Oracle comparisons inspect future actions and are not a deployable predictor; empirical engagement gains do not establish complete understanding of user interests.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Uses existing PinSage embeddings, Ward clustering and approximate nearest-neighbor search. A variable number of stored interests does not mean all are queried: the described retrieval samples three medoids. Oracle comparisons inspect future actions and are not a deployable predictor; empirical engagement gains do not establish complete understanding of user interests.

Contributors

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction