Predictive models capture expert uncertainty in process mining
Expert intuition is tacit. A popular way of tapping into that intuition is by using expert feedback. However, human expert feedback is sparse and non-trivial to obtain. In contrast, dark knowledge provides an incredibly accessible framework for capturing uncertainty. Considering this, it might be useful to train predictive models to forecast expert decisions. This would lead to a conceptual heatmap of possible actions an expert would take in a situation. Those models would resemble word suggestions when typing. The heatmap contains the expert intuition which the learner might seek to make their object of learning, but made visible.