Search papers, labs, and topics across Lattice.
The paper introduces Dual-view Calibration for Sequential Recommendation denoising (DC4SR), a framework that leverages the disagreement between a semantic prior (derived from a fine-tuned LLM) and a model-side posterior to denoise user interaction histories. This disagreement is used to iteratively refine both the semantic understanding and the model-side representations, aligning them with evolving user interests. Experiments show that DC4SR outperforms existing Transformer-based and LLM-based denoising methods, demonstrating improved robustness across training stages and noise conditions.
LLMs can denoise sequential recommendations by disagreeing with the recommendation model itself, leading to more robust performance against noisy user data.
Sequential recommendation seeks to model the evolution of user interests by capturing temporal user intent and item-level transition patterns. Transformer-based recommenders demonstrate a strong capacity for learning long-range and interpretable dependencies, yet remain vulnerable to behavioral noise that is misaligned with users'true preferences. Recent large language model (LLM)-based approaches attempt to denoise interaction histories through static semantic editing. Such methods neglect the learning dynamics of recommendation models and fail to account for the evolving nature of user interests. To address this limitation, we propose a Dual-view Calibration framework for Sequential Recommendation denoising (DC4SR). Specifically, we introduce a semantic prior, derived from an LLM fine-tuned via labeled historical interactions, to estimate the noise distribution from a semantic perspective. From the learning perspective, we further employ a model-side posterior that infers the noise distribution based on the model's learning dynamics. The disagreement between the two distributions is then leveraged to jointly refine semantic understanding and learning-aware model-side representations. Through iterative updates, dynamic dual-view calibration is achieved for both the global semantic prior and the model-side posterior, enabling consistent alignment with evolving user interests. Extensive experiments demonstrate that DC4SR consistently outperforms strong Transformer-based recommenders and LLM-based denoising methods, exhibiting enhanced robustness across training stages and noise conditions.
