When obtaining data mining results, decision makers can't know how these results were produced. Specifically, they don't know what input were these results based upon and how were they derived. So, I think explaining the mining processes will be a useful way to help decision makers verify mining results, just like many data mining platforms, such as SPSS Modeler, Rapidminer and so on. 

Althought this idea sounds reasonable, but I have not found any person who researches on it. So, I am very interested in discussing it with anyone.