Federated method finds communities in bipartite networks without sharing raw data
Federated Community Detection in Bipartite Networks
Machine Learning
Summary
Finding groups or communities in networks of two different types of things—like users and artists—usually requires combining all data in one place, which can be impossible due to privacy or size. The authors propose a method called FLICKER that works on pieces of the network stored separately, without sharing the raw data, by focusing on key directions in the data’s structure. They extend this to BiFLICKER for networks connecting two kinds of nodes, showing it can identify communities almost as well as centralized methods and even reveal interesting cross-country music listening groups. Their approach is mathematically supported and works well with small local datasets.
What this means in practice
- •For network analysts: Identify communities in bipartite networks distributed across multiple servers without centralizing sensitive data.
- •For music streaming platform engineers: Detect user-artist listener groups partitioned by geography while preserving data privacy for personalized recommendations.$Commercial implications: Enables privacy-conscious community detection in user-music interaction data beneficial for targeted recommendations.
Authors
Wanjie Wang
Abstract
We study federated spectral estimation when a large matrix is horizontally partitioned across servers that cannot share raw data. We propose FLICKER, a target-specific gradient-based power iteration that recovers a singular direction associated with a prescribed spectral target using only matrix-vector products on local blocks. Different target directions can be recovered independently and in parallel. We establish non-asymptotic convergence guarantees and provide a recursive grid-search procedure for spectral target selection. Building on FLICKER, we develop BiFLICKER for federated community detection in bipartite networks, together with a sequential projected variant for nearby or repeated singular directions. Under a sparse bipartite stochastic block model, we show that properly rescaled leading singular values from a reference server provide valid global targets, and that Sequential BiFLICKER achieves a vanishing misclustering rate and exact recovery under stronger sparsity conditions. Simulations show that BiFLICKER remains close to centralized spectral clustering and is more robust than existing federated spectral baselines when local sample sizes are small. On a LastFM user-artist network naturally partitioned by country, Sequential BiFLICKER attains an overall ARI of 0.94 against centralized clustering while revealing cross-country communities with distinct listening profiles.