RecThinker: An Agentic Framework for Tool-Augmented Reasoning in Recommendation

2026-03-10

AI Summary

This work addresses the limitations of existing recommender systems when faced with fragmented user profiles or sparse item metadata, stemming from their reliance on static pipelines and inability to actively acquire missing information. To overcome this, we propose RecThinker, a novel framework that introduces autonomous tool invocation and dynamic reasoning path planning into recommendation systems for the first time. Built upon an Analyze-Plan-Act architecture, RecThinker enables an agent to dynamically construct reasoning paths and proactively invoke specialized tools to enrich incomplete information. A self-augmented training strategy combining supervised fine-tuning and reinforcement learning significantly enhances both decision accuracy and tool utilization efficiency. Extensive experiments demonstrate that RecThinker consistently outperforms strong baselines across multiple benchmark datasets, validating its effectiveness and robustness in complex recommendation scenarios.

Technology Category

• Data Mining & Knowledge Management: Recommender Systems
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Reasoning under Uncertainty: Sequential Decision Making

Application Category

• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Economics and fairness of platforms and recommendation systems

Abstract

Large Language Models (LLMs) have revolutionized recommendation agents by providing superior reasoning and flexible decision-making capabilities. However, existing methods mainly follow a passive information acquisition paradigm, where agents either rely on static pre-defined workflows or perform reasoning with constrained information. It limits the agent's ability to identify information sufficiency, often leading to suboptimal recommendations when faced with fragmented user profiles or sparse item metadata. To address these limitations, we propose RecThinker, an agentic framework for tool-augmented reasoning in recommendation, which shifts recommendation from passive processing to autonomous investigation by dynamically planning reasoning paths and proactively acquiring essential information via autonomous tool-use. Specifically, RecThinker adopts an Analyze-Plan-Act paradigm, which first analyzes the sufficiency of user-item information and autonomously invokes tool-calling sequences to bridge information gaps between available knowledge and reasoning requirements. We develop a suite of specialized tools for RecThinker, enabling the model to acquire user-side, item-side, and collaborative information for better reasoning and user-item matching. Furthermore, we introduce a self-augmented training pipeline, comprising a Supervised Fine-Tuning (SFT) stage to internalize high-quality reasoning trajectories and a Reinforcement Learning (RL) stage to optimize for decision accuracy and tool-use efficiency. Extensive experiments on multiple benchmark datasets demonstrate that RecThinker consistently outperforms strong baselines in the recommendation scenario.

Problem

• recommendation
• information sufficiency
• tool-augmented reasoning
• large language models
• user-item matching

Innovation

• tool-augmented reasoning
• agentic recommendation
• autonomous tool-use
• Analyze-Plan-Act paradigm
• self-augmented training