Comparative results of K-nearest neighbor method on the data set consisting of process mining features and without process mining features