VoteGCL: Enhancing Graph-based Recommendations with Majority-Voting LLM-Rerank Augmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Nguyen, Minh-Anh, Nguyen, Bao, T., Ha Lan N., Hoang, Tuan Anh, Le, Duc-Trong, Le, Dung D.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915946577264640
author Nguyen, Minh-Anh
Nguyen, Bao
T., Ha Lan N.
Hoang, Tuan Anh
Le, Duc-Trong
Le, Dung D.
author_facet Nguyen, Minh-Anh
Nguyen, Bao
T., Ha Lan N.
Hoang, Tuan Anh
Le, Duc-Trong
Le, Dung D.
contents Recommendation systems often suffer from data sparsity caused by limited user-item interactions, which degrade their performance and amplify popularity bias in real-world scenarios. This paper proposes a novel data augmentation framework that leverages Large Language Models (LLMs) and item textual descriptions to enrich interaction data. By few-shot prompting LLMs multiple times to rerank items and aggregating the results via majority voting, we generate high-confidence synthetic user-item interactions, supported by theoretical guarantees based on the concentration of measure. To effectively leverage the augmented data in the context of a graph recommendation system, we integrate it into a graph contrastive learning framework to mitigate distributional shift and alleviate popularity bias. Extensive experiments show that our method improves accuracy and reduces popularity bias, outperforming strong baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2507_21563
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle VoteGCL: Enhancing Graph-based Recommendations with Majority-Voting LLM-Rerank Augmentation
Nguyen, Minh-Anh
Nguyen, Bao
T., Ha Lan N.
Hoang, Tuan Anh
Le, Duc-Trong
Le, Dung D.
Information Retrieval
Machine Learning
Recommendation systems often suffer from data sparsity caused by limited user-item interactions, which degrade their performance and amplify popularity bias in real-world scenarios. This paper proposes a novel data augmentation framework that leverages Large Language Models (LLMs) and item textual descriptions to enrich interaction data. By few-shot prompting LLMs multiple times to rerank items and aggregating the results via majority voting, we generate high-confidence synthetic user-item interactions, supported by theoretical guarantees based on the concentration of measure. To effectively leverage the augmented data in the context of a graph recommendation system, we integrate it into a graph contrastive learning framework to mitigate distributional shift and alleviate popularity bias. Extensive experiments show that our method improves accuracy and reduces popularity bias, outperforming strong baselines.
title VoteGCL: Enhancing Graph-based Recommendations with Majority-Voting LLM-Rerank Augmentation
topic Information Retrieval
Machine Learning
url https://arxiv.org/abs/2507.21563