Leveraging Business Process Improvement with Natural Language Processing and Organizational Semantic Knowledge

Contemporary organizations need to adapt their business processes swiftly to cope with ever-changing requirements. Requirement changes originate from a wide variety of sources. Business analysts gather these requirements, resolve conflicts, analyze impacts, and prepare actionable improvement plans. These tasks require a comprehensive knowledge of business processes and other entities within the organization. Business process model repositories, which may contain hundreds of models, are important sources of such cross-functional information. In this study, we introduce an approach which facilitates business process improvement by utilizing the comprehensive information covered by process models. Specifically, we associate requirements with other organizational entities based on their transitive relations with process models. To infer these associations, our approach makes use of natural language processing techniques and enterprise semantics. A quantitative evaluation of our approach, which took place within a major telecommunication company, displayed that it accurately detects associations between requirements and process models. Furthermore, semi-structured interviews with business analysts revealed that their expectations are high on efficiency increases due to the usage of this approach.