Explainable and Interpretable Forecasts on Non-Smooth Multivariate Time Series for Responsible Gameplay

2024-08-24

Knowledge Discovery and Data Mining

AI Summary

This paper addresses the problem of predicting excessive engagement risk from non-smooth multivariate time series (MTS) in online gaming scenarios. We propose the Actionable Forecasting Network (AFN), the first model designed for actionable forecasting in this setting. Unlike existing approaches that assume temporal smoothness, AFN jointly optimizes high-accuracy forecasting, generation of smooth and interpretable behavioral trajectories, and multi-dimensional feature attribution via an integrated architecture comprising attention mechanisms, a differentiable attribution module, and trajectory regularization. Evaluated on real-world gaming data, AFN reduces mean squared error (MSE) by 25%, doubles the proportion of high-risk players identified four weeks in advance (achieving 23%), and precisely pinpoints critical risk-inducing timesteps to inform timely interventions—resulting in an 18% reduction in subsequent excessive engagement behaviors.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Multiagent Systems: Adversarial Agents
• Machine Learning: Time-Series/Data Streams

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Multi-variate Time Series (MTS) forecasting has made large strides (with very negligible errors) through recent advancements in neural networks, e.g., Transformers. However, in critical situations like predicting gaming overindulgence that affects one's mental well-being; an accurate forecast without a contributing evidence (explanation) is irrelevant. Hence, it becomes important that the forecasts are Interpretable - intermediate representation of the forecasted trajectory is comprehensible; as well as Explainable - attentive input features and events are accessible for a personalized and timely intervention of players at risk. While the contributing state of the art research on interpretability primarily focuses on temporally-smooth single-process driven time series data, our online multi-player gameplay data demonstrates intractable temporal randomness due to intrinsic orthogonality between player's game outcome and their intent to engage further. We introduce a novel deep Actionable Forecasting Network (AFN), which addresses the inter-dependent challenges associated with three exclusive objectives - 1) forecasting accuracy; 2) smooth comprehensible trajectory and 3) explanations via multi-dimensional input features while tackling the challenges introduced by our non-smooth temporal data, together in one single solution. AFN establishes a it{new benchmark} via: (i) achieving 25% improvement on the MSE of the forecasts on player data in comparison to the SOM-VAE based SOTA networks; (ii) attributing unfavourable progression of a player's time series to a specific future time step(s), with the premise of eliminating near-future overindulgent player volume by over 18% with player specific actionable inputs feature(s) and (iii) proactively detecting over 23% (100% jump from SOTA) of the to-be overindulgent, players on an average, 4 weeks in advance.

Problem

• Interpretable forecasts for non-smooth multivariate gameplay data
• Explainable predictions to prevent gaming overindulgence
• Accurate forecasting with actionable player-specific interventions

Innovation

• Actionable Forecasting Network for non-smooth MTS
• Interpretable and explainable player behavior forecasts
• 25% MSE improvement over SOM-VAE benchmarks