Beyond Isolated Clients: Integrating Graph-Based Embeddings into Event Sequence Models

TL;DR

This paper introduces three strategies to incorporate user-item interaction graph structures into contrastive self-supervised learning models, improving prediction accuracy on large-scale event data.

Contribution

It presents novel, model-agnostic methods for integrating graph information into SSL for event sequence modeling, enhancing performance.

Findings

Improved accuracy up to 2.3% AUC on real datasets.

Graph density influences the effectiveness of integration strategies.

Structural information enhances user attribute prediction.

Abstract

Large-scale digital platforms generate billions of timestamped user-item interactions (events) that are crucial for predicting user attributes in, e.g., fraud prevention and recommendations. While self-supervised learning (SSL) effectively models the temporal order of events, it typically overlooks the global structure of the user-item interaction graph. To bridge this gap, we propose three model-agnostic strategies for integrating this structural information into contrastive SSL: enriching event embeddings, aligning client representations with graph embeddings, and adding a structural pretext task. Experiments on four financial and e-commerce datasets demonstrate that our approach consistently improves the accuracy (up to a 2.3% AUC) and reveals that graph density is a key factor in selecting the optimal integration strategy.