Gianmaria Tarantino, Data Scientist, Maps Group, Italy
Paul Cosma, Industrial PhD, KMD & DIKU
We present a case from a danish municipality using text processing, process mining and a process engine to support understanding the transformation from a non-process-aware electronic case and document management system to a process-aware case management system. We use a pipeline of tools consisting of text embeddings, regex patterns, and semantic similarity to match domain events identified jointly with domain experts with events extracted from free text data in the non-process-aware case and document management system. We apply the DisCoveR and Disco process miners to mine respectively declarative Dynamic Condition Response graphs and traditional, imperative transition graphs. We discuss the insights obtained and the benefits and drawbacks of using the different technologies in the presented case.