ThinkRec: Thinking-based recommendation via LLM

2025-05-21

AI Summary

Existing LLM4Rec approaches rely on superficial feature matching, yielding shallow recommendations with limited accuracy and interpretability. To address this, we propose DeepRec—the first System 2–driven deep reasoning framework explicitly designed for recommendation tasks, moving beyond intuitive (System 1) matching paradigms. Its core contributions are: (1) a *reasoning activation mechanism* that synthesizes reasoning chains and keyword-based summaries to explicitly model user behavioral logic within LLMs; and (2) an *instance-level expert fusion mechanism* that dynamically weights multiple domain-specific expert models via latent variables and integrates chain-of-thought (CoT) prompting for personalized inference. Evaluated on multiple real-world datasets, DeepRec achieves an average 12.7% improvement in NDCG@10 over state-of-the-art baselines. Moreover, it generates structured, verifiable reasoning paths, substantially enhancing both recommendation accuracy and interpretability.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Knowledge Representation and Reasoning: Reasoning with Beliefs

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• 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

Recent advances in large language models (LLMs) have enabled more semantic-aware recommendations through natural language generation. Existing LLM for recommendation (LLM4Rec) methods mostly operate in a System 1-like manner, relying on superficial features to match similar items based on click history, rather than reasoning through deeper behavioral logic. This often leads to superficial and erroneous recommendations. Motivated by this, we propose ThinkRec, a thinking-based framework that shifts LLM4Rec from System 1 to System 2 (rational system). Technically, ThinkRec introduces a thinking activation mechanism that augments item metadata with keyword summarization and injects synthetic reasoning traces, guiding the model to form interpretable reasoning chains that consist of analyzing interaction histories, identifying user preferences, and making decisions based on target items. On top of this, we propose an instance-wise expert fusion mechanism to reduce the reasoning difficulty. By dynamically assigning weights to expert models based on users' latent features, ThinkRec adapts its reasoning path to individual users, thereby enhancing precision and personalization. Extensive experiments on real-world datasets demonstrate that ThinkRec significantly improves the accuracy and interpretability of recommendations. Our implementations are available in anonymous Github: https://anonymous.4open.science/r/ThinkRec_LLM.

Problem

• Shifting LLM-based recommendations from superficial to reasoning-based approaches
• Enhancing recommendation accuracy via interpretable reasoning chains
• Improving personalization through dynamic expert model fusion

Innovation

• Thinking activation mechanism for reasoning chains
• Instance-wise expert fusion for personalization
• Synthetic reasoning traces for interpretable recommendations