Semantics-Aware Denoising: A PLM-Guided Sample Reweighting Strategy for Robust Recommendation

TL;DR

This paper introduces SAID, a semantic consistency-based reweighting strategy for implicit feedback in recommender systems, effectively reducing noise impact and improving prediction accuracy without complex training procedures.

Contribution

SAID is a novel, PLM-guided sample reweighting framework that enhances recommendation robustness by leveraging semantic similarity to identify noisy interactions.

Findings

Up to 2.2% relative AUC improvement over baselines

Effective noise reduction under high noise conditions

Simplifies denoising by only modifying the loss function

Abstract

Implicit feedback, such as user clicks, serves as the primary data source for modern recommender systems. However, click interactions inherently contain substantial noise, including accidental clicks, clickbait-induced interactions, and exploratory browsing behaviors that do not reflect genuine user preferences. Training recommendation models with such noisy positive samples leads to degraded prediction accuracy and unreliable recommendations. In this paper, we propose SAID (Semantics-Aware Implicit Denoising), a simple yet effective framework that leverages semantic consistency between user interests and item content to identify and downweight potentially noisy interactions. Our approach constructs textual user interest profiles from historical behaviors and computes semantic similarity with target item descriptions using pre-trained language model (PLM) based text encoders. The similarity scores are then transformed into sample weights that modulate the training loss, effectively reducing the impact of semantically inconsistent clicks. Unlike existing denoising methods that require complex auxiliary networks or multi-stage training procedures, SAID only modifies the loss function while keeping the backbone recommendation model unchanged. Extensive experiments on two real-world datasets demonstrate that SAID consistently improves recommendation performance, achieving up to 2.2% relative improvement in AUC over strong baselines, with particularly notable robustness under high noise conditions.