Opening

Process Mining at Scale

Bart Hompes (Artifex consultancy), Ekaterina Sabelnikova (Philips Industry Consulting) Process mining can be mistakenly positioned or interpreted as the “silver bullet” that will solve all problems, streamline processes, eliminate waste, and make operations efficient and productive. Though most academic research focuses on fancy new algorithms and supporting technologies, real world implementations face challenges that are more organizational than technological. In this talk we will discuss how to build a foundation for a successful implementation of process mining in large-scale organizations. Rather than focus on any specific process, we discuss the organizational aspect of setting up a process mining competency and its integration into the organization.

Using Text Processing to Extract Events for Mining a Non-process-aware Document Management System: A Case Study in a Danish Municipality

Paul Cosma (Copenhagen University and KMD) We  present  a  case  study  from  a  Danish  municipality  using text processing to construct an event log for a non-process-aware case  and  document  management  system  that  can  be  used  for  process analysis and mining. We use a pipeline of tools consisting of text embeddings, regex patterns, and semantic similarity to match domain events identified jointly with a domain expert with events extracted from free text data in the case and document management system. We apply the DisCoveR  process  miner  to  mine  declarative  Dynamic  Condition  Response graphs providing both a high-level description of the document management process and a domain-specific process from the refined eventlog of matched domain events.