Object-centric Process Mining (OCPM) is a new approach integrating objects into the process mining model. This offers a more realistic view of real word processes. Since OCPM is an upcoming topic in research and industry, we expect an increasing number of visualizations in the coming years. To evaluate the usability of these approaches for different process mining use cases, it is essential to have a taxonomy offering a structure about relevant visualization aspects. This paper presents a first proposal of such a taxonomy based on the what-why-how visualization framework. First, we developed the taxonomy based on visualization and interaction theory as well as existing OCPM tools. As the second step, we applied the taxonomy to two existing OCPM tools. The evaluation results show the usefulness of our taxonomy, covering relevant capabilities in existing tools and offering directions for future research. The existing tools offer valuable initial approaches to process discovery visualization, but several research directions remain open. For future work, we propose multiple directions on how to extend OCPM visualizations based on this taxonomy. This includes new visualization approaches, adding new process mining attributes, and improving the exploration capabilities of OCPM tools.

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

Process Visualizations for Object-Centric Process Mining: An Evaluation Taxonomy

  • Florian Eckhard,
  • Holger Wittges,
  • Stefanie Rinderle-Ma

摘要

Object-centric Process Mining (OCPM) is a new approach integrating objects into the process mining model. This offers a more realistic view of real word processes. Since OCPM is an upcoming topic in research and industry, we expect an increasing number of visualizations in the coming years. To evaluate the usability of these approaches for different process mining use cases, it is essential to have a taxonomy offering a structure about relevant visualization aspects. This paper presents a first proposal of such a taxonomy based on the what-why-how visualization framework. First, we developed the taxonomy based on visualization and interaction theory as well as existing OCPM tools. As the second step, we applied the taxonomy to two existing OCPM tools. The evaluation results show the usefulness of our taxonomy, covering relevant capabilities in existing tools and offering directions for future research. The existing tools offer valuable initial approaches to process discovery visualization, but several research directions remain open. For future work, we propose multiple directions on how to extend OCPM visualizations based on this taxonomy. This includes new visualization approaches, adding new process mining attributes, and improving the exploration capabilities of OCPM tools.