ProReco: A Process Discovery Recommender System

ProReco addresses a real decision-making challenge: The paper identifies that selecting among multiple process discovery algorithms is "time-consuming and error-prone" when done manually, demonstrating how AI systems can reduce human burden in technical decision-making tasks.

Transparency is built into the recommendation system: ProReco incorporates "eXplainable AI (XAI) techniques to provide explanations for its recommendations," meaning users can understand why the system suggests a particular algorithm rather than receiving a black-box recommendation.

The system accounts for diverse user needs: Rather than assuming one algorithm fits all situations, ProReco considers both "user preferences and event log characteristics," recognizing that different stakeholders may have different priorities and that context matters in algorithmic selection.

Responsible AI requires acknowledging competing values: The paper notes there is "no consensus on a dominating discovery algorithm" due to "competing quality measures," highlighting that responsible AI systems must navigate trade-offs between different success criteria rather than pretending one solution is universally optimal.

Explainability enables better human oversight: By providing explanations alongside recommendations, ProReco supports human decision-makers in validating whether the system's suggestions align with their actual needs, rather than passively accepting automated choices.