Scholarlink

<p>Single-pattern complex event detection technology efficiently extracts valuable information from event streams by leveraging relationships among event attributes, detection rules, and algebraic operations. However, existing methods face significant challenges in achieving shared detection for multiple complex event queries that involve different operators but share identical operands in high-density streams. This difficulty arises from the semantic diversity of the operators and the lack of built-in sharing mechanisms in current approaches. To address this limitation, this paper presents a multi-pattern complex event detection method based on pattern rewrite for high-density event streams. The proposed approach proceeds as follows. First, we analyze the structural and semantic characteristics of complex event detection expressions. Second, we identify the inherent transformational relationships among these expressions. Third, based on these relationships, we establish a set of pattern rewrite rules and properties. Fourth, we apply these rules to rewrite diverse complex event patterns, thereby generating shareable sub-patterns. Finally, these identical sub-patterns are merged to enable shared processing across multiple distinct complex event queries, significantly reducing redundancy in storage, computation, and result generation. Experimental results demonstrate that the proposed method outperforms existing general-purpose complex event detection approaches for high-density event streams.</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

Leveraging pattern rewriting for multi-pattern complex event detection in high-density streams

  • 期刊论文

  • 发表日期：

    2026年6月1日

查看全文
  • Jianhua Wang,
  • Yi Lin Zhang,
  • Qianming Chen

摘要

Single-pattern complex event detection technology efficiently extracts valuable information from event streams by leveraging relationships among event attributes, detection rules, and algebraic operations. However, existing methods face significant challenges in achieving shared detection for multiple complex event queries that involve different operators but share identical operands in high-density streams. This difficulty arises from the semantic diversity of the operators and the lack of built-in sharing mechanisms in current approaches. To address this limitation, this paper presents a multi-pattern complex event detection method based on pattern rewrite for high-density event streams. The proposed approach proceeds as follows. First, we analyze the structural and semantic characteristics of complex event detection expressions. Second, we identify the inherent transformational relationships among these expressions. Third, based on these relationships, we establish a set of pattern rewrite rules and properties. Fourth, we apply these rules to rewrite diverse complex event patterns, thereby generating shareable sub-patterns. Finally, these identical sub-patterns are merged to enable shared processing across multiple distinct complex event queries, significantly reducing redundancy in storage, computation, and result generation. Experimental results demonstrate that the proposed method outperforms existing general-purpose complex event detection approaches for high-density event streams.