Scholarlink

<p>In recent years, the problem of mining frequent weighted utility patterns (FWUP), a variation of frequent patterns, has garnered significant research interest. However, traditional approaches often generate a large number of patterns and typically assume static item weights. In real-world applications, item weights are influenced by various factors and may fluctuate over time. To address these issues, this paper introduces the problem of mining frequent weighted utility closed patterns (FWUCP), a concise and lossless representation of FWUPs, from dynamic quantitative databases where item weights are adjustable. We propose FWUCWP, an efficient algorithm that leverages the WUNList structure and integrates three pruning strategies, WUNList subset relation pruning, maximal twu combination pruning, and enclosure-based pruning, to reduce the search space. Experimental evaluations conducted on six real-world datasets, characterized by diverse densities and scales reaching up to 1.1 million transactions and over 46,000 distinct items, demonstrate that FWUCWP is up to 30 times faster than the state-of-the-art algorithm for mining FWUCP on sparse datasets, while also maintaining competitive memory usage.</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

Efficiently mining frequent weighted utility closed patterns with pruning strategies from dynamic quantitative databases

  • 期刊论文

  • 发表日期：

    2026年2月5日

查看全文
  • Nguyen Le,
  • Ham Nguyen,
  • Minh Nguyen,
  • Huong Bui,
  • Unil Yun,
  • Bay Vo

摘要

In recent years, the problem of mining frequent weighted utility patterns (FWUP), a variation of frequent patterns, has garnered significant research interest. However, traditional approaches often generate a large number of patterns and typically assume static item weights. In real-world applications, item weights are influenced by various factors and may fluctuate over time. To address these issues, this paper introduces the problem of mining frequent weighted utility closed patterns (FWUCP), a concise and lossless representation of FWUPs, from dynamic quantitative databases where item weights are adjustable. We propose FWUCWP, an efficient algorithm that leverages the WUNList structure and integrates three pruning strategies, WUNList subset relation pruning, maximal twu combination pruning, and enclosure-based pruning, to reduce the search space. Experimental evaluations conducted on six real-world datasets, characterized by diverse densities and scales reaching up to 1.1 million transactions and over 46,000 distinct items, demonstrate that FWUCWP is up to 30 times faster than the state-of-the-art algorithm for mining FWUCP on sparse datasets, while also maintaining competitive memory usage.