ConsRec: Denoising Sequential Recommendation through User-Consistent Preference Modeling

2025-05-28

AI Summary

To address performance degradation in recommendation systems caused by noisy user interactions—such as accidental clicks or non-preferential behaviors—this paper proposes a graph connectivity-constrained interaction denoising method. We construct a user-item interaction graph, incorporate BERT-based textual embeddings to model item semantic similarity, and extract the maximum connected subgraph to explicitly identify and remove structurally inconsistent interactions. This work is the first to jointly model user preference consistency and graph connectivity constraints, enabling semantic-robust sequence denoising at the topological level. Evaluated on Yelp and Amazon datasets, the proposed method improves recommendation accuracy by 13% over state-of-the-art baselines, while significantly enhancing target-item relevance and sequential semantic compactness.

Technology Category

• Data Mining & Knowledge Management: Recommender Systems
• Machine Learning: Learning Preferences or Rankings
• Knowledge Representation and Reasoning: Preferences

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

User-item interaction histories are pivotal for sequential recommendation systems but often include noise, such as unintended clicks or actions that fail to reflect genuine user preferences. To address this issue, we propose the User-Consistent Preference-based Sequential Recommendation System (ConsRec), designed to capture stable user preferences and filter noisy items from interaction histories. Specifically, ConsRec constructs a user-interacted item graph, learns item similarities from their text representations, and then extracts the maximum connected subgraph from the user-interacted item graph for denoising items. Experimental results on the Yelp and Amazon Product datasets illustrate that ConsRec achieves a 13% improvement over baseline recommendation models, showing its effectiveness in denoising user-interacted items. Further analysis reveals that the denoised interaction histories form semantically tighter clusters of user-preferred items, leading to higher relevance scores for ground-truth targets and more accurate recommendations. All codes are available at https://github.com/NEUIR/ConsRec.

Problem

• Denoising user-item interaction histories for sequential recommendation
• Capturing stable user preferences by filtering noisy items
• Improving recommendation accuracy through semantically clustered user preferences

Innovation

• Constructs user-interacted item graph
• Learns item similarities from text
• Extracts maximum connected subgraph for denoising