Querying Workflow Logs

• Main Authors: Yan Tang, Isaac Mackey, Jianwen Su
• Format: Article
• Language: English
• Published: MDPI AG 2018-01-01
• Series: Information
• Subjects: business analytics; workflow logs; query languages
• Online Access: http://www.mdpi.com/2078-2489/9/2/25

A business process or workflow is an assembly of tasks that accomplishes a business goal. Business process management is the study of the design, configuration/implementation, enactment and monitoring, analysis, and re-design of workflows. The traditional methodology for the re-design and improvement of workflows relies on the well-known sequence of extract, transform, and load (ETL), data/process warehousing, and online analytical processing (OLAP) tools. In this paper, we study the ad hoc queryiny of process enactments for (data-centric) business processes, bypassing the traditional methodology for more flexibility in querying. We develop an algebraic query language based on “incident patterns” with four operators inspired from Business Process Model and Notation (BPMN) representation, allowing the user to formulate ad hoc queries directly over workflow logs. A formal semantics of this query language, a preliminary query evaluation algorithm, and a group of elementary properties of the operators are provided.