Leveraging Scene Context with Dual Networks for Sequential User Behavior Modeling

2025-09-30

AI Summary

Existing methods for user behavior sequence prediction often neglect or coarsely model contextual scene features (e.g., app sub-interfaces), limiting accurate interest modeling. Method: We propose the Dual Sequence Prediction Network (DSPnet), the first model to explicitly capture users’ dynamic interests across both item and scene dimensions, along with their co-evolutionary relationships. DSPnet employs a dual-stream architecture, a sequence feature enhancement module, and a novel Conditional Contrastive Regularization (CCR) loss to improve sequential generalization. We provide theoretical justification for its joint relational modeling capability. Contribution/Results: DSPnet achieves statistically significant improvements over state-of-the-art baselines on multiple public and industrial datasets. After online deployment, it yields a +0.04 percentage point lift in CTR, and increases order volume and GMV by 0.78% and 0.64%, respectively.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Computer Vision: Scene Analysis & Understanding
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Modeling sequential user behaviors for future behavior prediction is crucial in improving user's information retrieval experience. Recent studies highlight the importance of incorporating contextual information to enhance prediction performance. One crucial but usually neglected contextual information is the scene feature which we define as sub-interfaces within an app, created by developers to provide specific functionalities, such as ``text2product search" and ``live" modules in e-commence apps. Different scenes exhibit distinct functionalities and usage habits, leading to significant distribution gap in user engagement across them. Popular sequential behavior models either ignore the scene feature or merely use it as attribute embeddings, which cannot effectively capture the dynamic interests and interplay between scenes and items when modeling user sequences. In this work, we propose a novel Dual Sequence Prediction networks (DSPnet) to effectively capture the dynamic interests and interplay between scenes and items for future behavior prediction. DSPnet consists of two parallel networks dedicated to learn users' dynamic interests over items and scenes, and a sequence feature enhancement module to capture the interplay for enhanced future behavior prediction. Further, we introduce a Conditional Contrastive Regularization (CCR) loss to capture the invariance of similar historical sequences. Theoretical analysis suggests that DSPnet is a principled way to learn the joint relationships between scene and item sequences. Extensive experiments are conducted on one public benchmark and two collected industrial datasets. The method has been deployed online in our system, bringing a 0.04 point increase in CTR, 0.78% growth in deals, and 0.64% rise in GMV. The codes are available at this anonymous github: extcolor{blue}{https://anonymous.4open.science/r/DSPNet-ForPublish-2506/}.

Problem

• Modeling sequential user behaviors for future behavior prediction
• Incorporating scene context to enhance prediction performance
• Capturing dynamic interests and interplay between scenes and items

Innovation

• Dual networks model scene and item sequences
• Sequence feature enhancement captures scene-item interplay
• Contrastive regularization preserves historical sequence invariance