Database Paper Browser

Back to papers

MultiBiSage: A Web-Scale Recommendation System Using Multiple Bipartite Graphs at Pinterest

Summary: Breaks Pinterest’s heterogeneous entity interactions into multiple disjoint bipartite graphs and introduces MultiBiSage, a data-efficient GCN that fuses embeddings from these bipartite graphs while requiring minimal changes to existing PinSage infrastructure. Trained on six bipartite graphs at web scale, MultiBiSage significantly outperforms deployed PinSage on engagement metrics and generalizes to public datasets. (summarized by gpt-5-mini on Feb 09 2026)

Paper ID
13331
Venue
VLDB
Year
2023
Pagerank
4.5990229e-05
Overall Rank
8,045 | 44.04%
DOI
10.14778/3574245.3574262

Incoming Non-self Citations Over Time

Authors

Incoming Citations (Sorted by Pagerank)

Showing 2 of 2 citing papers.

Previous Page 1 / 1 Next

Outgoing Citations (Sorted by Pagerank)

Showing 0 of 0 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank Cited Paper Year Venue Pagerank
Previous Page 1 / 1 Next

Semantically Similar Papers