Generative Chain of Behavior for User Trajectory Prediction

2026-01-26

AI Summary

Existing sequential recommendation systems primarily focus on single-step next-item prediction, struggling to capture the dependencies among multi-step user behaviors and the evolution of long-term preferences. This work proposes the Generative Behavior Chain (GCB) framework, which introduces a unified autoregressive generative model to represent user interaction sequences as semantic behavior chains. GCB constructs a structured semantic space by combining RQ-VAE with k-means clustering to generate discrete semantic IDs, and leverages a Transformer-based architecture for conditional autoregressive generation of multi-step trajectories. Experimental results demonstrate that GCB significantly outperforms state-of-the-art methods across multiple benchmark datasets, achieving superior performance in both multi-step prediction accuracy and trajectory consistency.

Technology Category

• Natural Language Processing: Generation
• Reasoning under Uncertainty: Sequential Decision Making
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• User Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Modeling long-term user behavior trajectories is essential for understanding evolving preferences and enabling proactive recommendations. However, most sequential recommenders focus on next-item prediction, overlooking dependencies across multiple future actions. We propose Generative Chain of Behavior (GCB), a generative framework that models user interactions as an autoregressive chain of semantic behaviors over multiple future steps. GCB first encodes items into semantic IDs via RQ-VAE with k-means refinement, forming a discrete latent space that preserves semantic proximity. On top of this space, a transformer-based autoregressive generator predicts multi-step future behaviors conditioned on user history, capturing long-horizon intent transitions and generating coherent trajectories. Experiments on benchmark datasets show that GCB consistently outperforms state-of-the-art sequential recommenders in multi-step accuracy and trajectory consistency. Beyond these gains, GCB offers a unified generative formulation for capturing user preference evolution.

Problem

• user trajectory prediction
• long-term behavior modeling
• sequential recommendation
• multi-step prediction
• preference evolution

Innovation

• Generative Chain of Behavior
• RQ-VAE
• semantic ID
• autoregressive trajectory prediction
• sequential recommendation