Scholarlink

<p>Recommender systems (RSs) still struggle to effectively model complex, high-dimensional user-item interactions and incorporate contextual information. We propose SimGOL (Similarity-based Graph Optimization and Learning), a graph-based framework that addresses these challenges by modeling user and item interactions as separate graphs via Graph Convolutional Networks (GCNs) and integrating them through a Flexible Fusion Mechanism (FFM). This approach captures similarity-based connections in each graph and uses the FFM’s attention and gating to dynamically fuse the user and item representations. We evaluated our proposed model on six benchmark datasets, ranging from 82,790 to 1,000,209 interactions, comparing its performance against eight state-of-the-art baseline models. SimGOL achieved up to a 4.6% higher average AUC and a 14.3% lower RMSE than the best-performing baseline models across these datasets, demonstrating significantly improved recommendation accuracy. These results highlight SimGOL’s enhanced adaptability and interpretability, underscoring the framework’s robustness and scalability across diverse recommendation scenarios.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

SimGOL: a similarity-based graph optimization and learning framework with flexible fusion for recommender systems

  • 期刊论文

  • 发表日期：

    2026年5月21日

查看全文
  • Sang-Woong Lee,
  • Saqib Ali,
  • Amir Masoud Rahmani,
  • Gholamreza Zare,
  • Pegah Malekpour Alamdari,
  • Parisa Khoshvaght,
  • Mehdi Hosseinzadeh

摘要

Recommender systems (RSs) still struggle to effectively model complex, high-dimensional user-item interactions and incorporate contextual information. We propose SimGOL (Similarity-based Graph Optimization and Learning), a graph-based framework that addresses these challenges by modeling user and item interactions as separate graphs via Graph Convolutional Networks (GCNs) and integrating them through a Flexible Fusion Mechanism (FFM). This approach captures similarity-based connections in each graph and uses the FFM’s attention and gating to dynamically fuse the user and item representations. We evaluated our proposed model on six benchmark datasets, ranging from 82,790 to 1,000,209 interactions, comparing its performance against eight state-of-the-art baseline models. SimGOL achieved up to a 4.6% higher average AUC and a 14.3% lower RMSE than the best-performing baseline models across these datasets, demonstrating significantly improved recommendation accuracy. These results highlight SimGOL’s enhanced adaptability and interpretability, underscoring the framework’s robustness and scalability across diverse recommendation scenarios.