Scholarlink

Link prediction in social networks is a critical problem in Social Network Analysis, essential for uncovering hidden connections and enhancing user engagement on platforms. This paper introduces a novel Deep Reinforcement Learning (DRL) approach to improve link prediction accuracy by adaptively selecting optimal similarity metrics. Unlike traditional approaches with fixed measures, our method leverages structural and temporal data from social networks, dynamically adjusting selections to align with the network’s unique characteristics. Experimental results across various real-world networks suggest that the proposed method achieves higher accuracy than other state-of-the-art similarity-based link prediction methods, confirming the approach’s robustness and adaptability.

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

Enhanced Link Prediction in Social Networks Leveraging Reinforcement Learning and Similarity Algorithms

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Bouchra Bouchoul,
  • Ibtihel Rezaiguia,
  • Fatima Benbouzid-Si Tayeb

摘要

Link prediction in social networks is a critical problem in Social Network Analysis, essential for uncovering hidden connections and enhancing user engagement on platforms. This paper introduces a novel Deep Reinforcement Learning (DRL) approach to improve link prediction accuracy by adaptively selecting optimal similarity metrics. Unlike traditional approaches with fixed measures, our method leverages structural and temporal data from social networks, dynamically adjusting selections to align with the network’s unique characteristics. Experimental results across various real-world networks suggest that the proposed method achieves higher accuracy than other state-of-the-art similarity-based link prediction methods, confirming the approach’s robustness and adaptability.