Scholarlink

Topic-Sensitive PageRank (TSPR) is a widely used algorithm in recommender systems and machine learning. However, the TSPR query is time-consuming as it requires multiple rounds of iterative computation. To accelerate the TSPR query, we propose an efficient TSPR query algorithm, \(\textsf{FasTSPR}\) , that accelerates the TSPR query by exploiting the previous TSPR query. Specifically, \(\textsf{FasTSPR}\) borrows the computation result of the previous TSPR query when performing the current TSPR query, avoiding redundant calculations and thus accelerating the TSPR query.

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

Accelerating Topic-Sensitive PageRank by Exploiting the Query History

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Shufeng Gong,
  • Zhixin Zhang,
  • Jing Lu,
  • Yanfeng Zhang,
  • Cong Fu,
  • Ge Yu

摘要

Topic-Sensitive PageRank (TSPR) is a widely used algorithm in recommender systems and machine learning. However, the TSPR query is time-consuming as it requires multiple rounds of iterative computation. To accelerate the TSPR query, we propose an efficient TSPR query algorithm, \(\textsf{FasTSPR}\) , that accelerates the TSPR query by exploiting the previous TSPR query. Specifically, \(\textsf{FasTSPR}\) borrows the computation result of the previous TSPR query when performing the current TSPR query, avoiding redundant calculations and thus accelerating the TSPR query.