This looks interesting because it add the ability to put the user in the loop of fixing/annotating the problematic observations relatively easily. I like the ex

This looks interesting because it add the ability to put the user in the loop of fixing/annotating the problematic observations relatively easily. I like the example of Tinder for data. Are the examples picked those that have the highest objective function error rate, or something similar? Does this apply only to text classification problems? Are there examples where this could be applied to tabular data?