Process-to-Text: A Framework for the Automatic Generation of Natural Language Descriptions of Processes

Abstract

Effectively presenting and comprehending processes is a challenging task. This thesis introduces different solutions for communicating process knowledge via natural language. We propose a taxonomy of relevant process descriptions based on a fuzzy protoform model, validated by medical experts. We introduce the Process-to-Text framework, building on the Data- To-Text architecture, leverages process mining techniques in an ontology-driven system using fuzzy logic, for the generation of accurate and contextually-aware process descriptions. Additionally, we introduce C-4PM, a conversational agent for declarative process mining, enhancing interaction and knowledge inference through natural language. This work signifies a stride towards making process mining accessible and understandable for a broader audience.