为对话式AI设计可动态响应的流程建模标准扩展
BPMN4CAI: A BPMN Extension for Modeling Dynamic Conversational AI
- 在BPMN基础上扩展新元素，支持上下文感知交互
- 案例验证其能提升AI流程的自适应决策与透明度
- 适合企业流程设计者与AI系统开发者参考
对话式AI系统（如聊天机器人、虚拟助手）在数字业务流程中日益重要，但现有业务流程建模与标注（BPMN）标准在表达动态、上下文敏感交互时存在局限。本文通过设计科学方法，提出一种符合BPMN规范的扩展（BPMN4CAI），系统性地拓展了原有元素并引入专用组件。通过案例研究验证该框架的适用性与有效性，结果表明，BPMN4CAI能有效支持对话式AI在业务流程中的自适应决策、稳健的上下文管理及透明交互。
原文摘要 · Abstract (English)
Conversational AI systems, such as chatbots and virtual assistants, are becoming increasingly important to digital business processes. However, the established Business Process Model and Notation (BPMN) standard faces challenges when representing dynamic, context-sensitive interactions. This paper addresses this methodological and practical research gap by developing a standard-compliant BPMN extension (BPMN4CAI). Using Design Science Research methodology, this paper develops an approach that systematically extends existing BPMN elements and incorporates specialized components. The applicability and relevance of the BPMN4CAI framework are demonstrated and evaluated through a case study. The results show that the BPMN4CAI extension facilitates adaptive decision-making processes, robust context management, and transparent interactions for Conversational AI within business processes.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。