STGIN: Spatial-Temporal Graph Interaction Network for Large-scale POI Recommendation

TL;DR

This paper introduces STGIN, a novel spatial-temporal graph interaction network for large-scale POI recommendation, which captures diverse user interests across different contexts and adapts efficiently to real-time changes, outperforming existing models.

Contribution

The paper proposes a new graph model that constructs multiple subgraphs for different spatial-temporal contexts and an industry-friendly framework for real-time interest tracking, addressing key limitations of prior methods.

Findings

Outperforms state-of-the-art models on real-world datasets.

Deployed in a large e-commerce platform, improving CTR by 1.1%.

Increases RPM by 6.3%.

Abstract

In Location-Based Services, Point-Of-Interest(POI) recommendation plays a crucial role in both user experience and business opportunities. Graph neural networks have been proven effective in providing personalized POI recommendation services. However, there are still two critical challenges. First, existing graph models attempt to capture users' diversified interests through a unified graph, which limits their ability to express interests in various spatial-temporal contexts. Second, the efficiency limitations of graph construction and graph sampling in large-scale systems make it difficult to adapt quickly to new real-time interests. To tackle the above challenges, we propose a novel Spatial-Temporal Graph Interaction Network. Specifically, we construct subgraphs of spatial, temporal, spatial-temporal, and global views respectively to precisely characterize the user's interests in various contexts. In addition, we design an industry-friendly framework to track the user's latest interests. Extensive experiments on the real-world dataset show that our method outperforms state-of-the-art models. This work has been successfully deployed in a large e-commerce platform, delivering a 1.1% CTR and 6.3% RPM improvement.