Knowing your FATE: Friendship, Action and Temporal Explanations for User Engagement Prediction on Social Apps

TL;DR

This paper introduces FATE, an explainable neural framework for user engagement prediction on social apps, integrating friendships, actions, and temporal factors to improve accuracy and interpretability.

Contribution

The paper proposes a novel, explainable neural model combining graph neural networks and attention mechanisms for user engagement prediction in social apps.

Findings

FATE outperforms state-of-the-art methods by approximately 10% in error reduction.

FATE achieves about 20% faster training and inference times.

The explanations generated by FATE are quantitatively and qualitatively effective.

Abstract

With the rapid growth and prevalence of social network applications (Apps) in recent years, understanding user engagement has become increasingly important, to provide useful insights for future App design and development. While several promising neural modeling approaches were recently pioneered for accurate user engagement prediction, their black-box designs are unfortunately limited in model explainability. In this paper, we study a novel problem of explainable user engagement prediction for social network Apps. First, we propose a flexible definition of user engagement for various business scenarios, based on future metric expectations. Next, we design an end-to-end neural framework, FATE, which incorporates three key factors that we identify to influence user engagement, namely friendships, user actions, and temporal dynamics to achieve explainable engagement predictions. FATE is based on a tensor-based graph neural network (GNN), LSTM and a mixture attention mechanism, which allows for (a) predictive explanations based on learned weights across different feature categories, (b) reduced network complexity, and (c) improved performance in both prediction accuracy and training/inference time. We conduct extensive experiments on two large-scale datasets from Snapchat, where FATE outperforms state-of-the-art approaches by ${\approx}10\%$ error and ${\approx}20\%$ runtime reduction. We also evaluate explanations from FATE, showing strong quantitative and qualitative performance.