Scholarlink

<p>Analyzing business process change impacts in uncertain environments is critical for ensuring the reliable management of business systems. Existing methods primarily analyze dependencies between control flow and data flow within processes, often overlooking how external environmental changes affect execution. Meanwhile, a large amount of data and decisions are embedded in processes, and changes in the environment or data increase the complexity of process management and maintenance. To address these limitations, we propose a context-aware method for analyzing and propagating change impacts. First, we construct a layered “Process–Service–Decision” architecture. By linking context data to Decision Model and Notation (DMN) inputs via decision service interfaces, we achieve unified modeling of internal processes and external environments. For internal process data, we construct a data dependency graph to identify how data changes affect subsequent process execution. For external environmental data, message boundary events are used to trigger the re-evaluation of incremental decisions. We then analyze how collaborative changes in context data propagate and impact the integrated model, ensuring correct decision invocation and maintaining consistency. The method employs a parallelizable dependency propagation algorithm that can efficiently process large-scale, high-concurrency sensor data streams. Finally, experiments implemented in Python on both synthetic and real-world datasets show that our method outperforms several baselines in accuracy, efficiency, and robustness for change impact identification.</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

Context-aware change impact analysis for integrated process–decision models in ubiquitous environments

  • 期刊论文

  • 发表日期：

    2026年4月24日

查看全文
  • Huijing Hao,
  • Xianwen Fang,
  • Daoyu Kan,
  • Chenliang Hao

摘要

Analyzing business process change impacts in uncertain environments is critical for ensuring the reliable management of business systems. Existing methods primarily analyze dependencies between control flow and data flow within processes, often overlooking how external environmental changes affect execution. Meanwhile, a large amount of data and decisions are embedded in processes, and changes in the environment or data increase the complexity of process management and maintenance. To address these limitations, we propose a context-aware method for analyzing and propagating change impacts. First, we construct a layered “Process–Service–Decision” architecture. By linking context data to Decision Model and Notation (DMN) inputs via decision service interfaces, we achieve unified modeling of internal processes and external environments. For internal process data, we construct a data dependency graph to identify how data changes affect subsequent process execution. For external environmental data, message boundary events are used to trigger the re-evaluation of incremental decisions. We then analyze how collaborative changes in context data propagate and impact the integrated model, ensuring correct decision invocation and maintaining consistency. The method employs a parallelizable dependency propagation algorithm that can efficiently process large-scale, high-concurrency sensor data streams. Finally, experiments implemented in Python on both synthetic and real-world datasets show that our method outperforms several baselines in accuracy, efficiency, and robustness for change impact identification.