Organizational Coupling as a Leading Indicator of Microservice Architecture Degradation

Title:Artifact Package for Organizational Coupling as a Leading Indicator of Microservice Architecture Degradation “Organizational Coupling → Service Degradation” Study Description:This artifact package enables full reproduction of the empirical results reported in the paper “Organizational Coupling → Service Degradation”. The artifact implements a three-stage analysis pipeline: Change Impact Analyzer (CIMET2) – extracts architectural degradation metrics (e.g., stdLOMLC, maxSSIC) from repository history. OC Computation Module – computes weekly Organizational Coupling (OC) values from version control and service mappings. Analytic Processing Pipeline – performs causality testing, recurrence analysis, robustness checks, and early-warning modeling. The package includes: All source code required to reproduce the analyses Step-by-step documentation (README and INSTALL guides) Pre-produced intermediate and final results under the Data/ directory A sample combined dataset used as analytic input The artifact reproduces: Granger causality results Prewhitened cross-correlation analyses Early-warning performance metrics (AUC-PR, precision@k) Recurrence pattern outputs Robustness and transformation analyses All outputs are generated as CSV artifacts and stored in structured data folders. The package is self-contained and does not require NDA-restricted data.### 标题：《作为微服务架构退化先行指标的组织耦合度》配套工件包 “组织耦合度→服务退化”研究 本配套工件包可完整复现论文《组织耦合度→服务退化》中报告的实证研究结果。该工件包实现了三阶段分析流水线： 1. 变更影响分析器（Change Impact Analyzer, CIMET2）：从代码仓库历史中提取架构退化指标（如stdLOMLC、maxSSIC）。 2. 组织耦合度（Organizational Coupling, OC）计算模块：从版本控制系统与服务映射信息中计算周度组织耦合度数值。 3. 分析处理流水线：执行因果检验、复发分析、稳健性检验与预警建模工作。 本工件包包含以下内容： - 复现所有分析所需的全部源代码 - 分步操作文档（含README与安装指南） - Data/目录下预先生成的中间结果与最终结果 - 用于分析输入的组合样本数据集 本工件包可复现以下结果： - 格兰杰因果检验结果 - 预白化交叉相关分析结果 - 预警性能指标（AUC-PR、precision@k） - 复发模式输出结果 - 稳健性与变换分析结果 所有输出结果均以CSV工件格式生成，并存储于结构化数据文件夹中。本工件包为独立完整的包，无需受保密协议（NDA）限制的数据。