Scholarlink

Community-based recommender systems have attracted substantial research interest by grouping individuals to leverage collective data and identify group traits. Previous methods focused on community detection but neglected the impact of users’ social activities and interests on preferences. Our Community-Based Product Recommendation System (CBPRS) uses social data to form communities based on past interactions within the target domain, enabling personalized suggestions. Evaluated using the Amazon dataset, CBPRS utilizes a network graph of product purchases to simulate a network of interests. Communities identified from frequently purchased items provide a basis for diverse recommendations. Among various algorithms, our findings highlight the Louvain Method as the most effective for our large-scale dataset, outperforming other community detection methods like Girvan-Newman.

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

Community Detection Based Recommender System Framework Using State of the Art Algorithms

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Chaitali Choudhary,
  • Inder Singh,
  • Manoj Kumar

摘要

Community-based recommender systems have attracted substantial research interest by grouping individuals to leverage collective data and identify group traits. Previous methods focused on community detection but neglected the impact of users’ social activities and interests on preferences. Our Community-Based Product Recommendation System (CBPRS) uses social data to form communities based on past interactions within the target domain, enabling personalized suggestions. Evaluated using the Amazon dataset, CBPRS utilizes a network graph of product purchases to simulate a network of interests. Communities identified from frequently purchased items provide a basis for diverse recommendations. Among various algorithms, our findings highlight the Louvain Method as the most effective for our large-scale dataset, outperforming other community detection methods like Girvan-Newman.