RadarSeq: A Temporal Vision Framework for User Churn Prediction via Radar Chart Sequences

2025-06-18

AI Summary

To address the challenge of predicting implicit, unlabeled, and dynamically evolving user churn in non-subscription-based gig platforms, this paper proposes a radar-chart sequence modeling framework with spatiotemporal joint learning. We pioneer the encoding of daily multidimensional behavioral features into radar-chart image sequences—overcoming the temporal information loss inherent in conventional snapshot-based representations. A lightweight and interpretable CV-LSTM architecture is designed, integrating a pretrained CNN encoder with a bidirectional LSTM and an integrated visualization-based attribution module. Evaluated on a large-scale real-world dataset, our method achieves absolute improvements of 17.7%, 29.4%, and 16.1% in F1-score, precision, and AUC, respectively, significantly outperforming traditional models and ViT-based baselines. This work establishes an efficient, production-ready paradigm for early identification of implicit churn.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Predicting user churn in non-subscription gig platforms, where disengagement is implicit, poses unique challenges due to the absence of explicit labels and the dynamic nature of user behavior. Existing methods often rely on aggregated snapshots or static visual representations, which obscure temporal cues critical for early detection. In this work, we propose a temporally-aware computer vision framework that models user behavioral patterns as a sequence of radar chart images, each encoding day-level behavioral features. By integrating a pretrained CNN encoder with a bidirectional LSTM, our architecture captures both spatial and temporal patterns underlying churn behavior. Extensive experiments on a large real-world dataset demonstrate that our method outperforms classical models and ViT-based radar chart baselines, yielding gains of 17.7 in F1 score, 29.4 in precision, and 16.1 in AUC, along with improved interpretability. The framework's modular design, explainability tools, and efficient deployment characteristics make it suitable for large-scale churn modeling in dynamic gig-economy platforms.

Problem

• Predicting implicit user churn in gig platforms without explicit labels
• Overcoming limitations of static methods by capturing temporal behavioral cues
• Improving accuracy and interpretability in dynamic user behavior modeling

Innovation

• Temporal vision framework for churn prediction
• Radar chart sequences encode behavior patterns
• CNN-LSTM hybrid captures spatiotemporal features