I agree that it makes more sense to be comparing the hypernym_paths () rather than the min_depth () for the synsets. I suppose it could get the lowest hypernyms for each possible path intersection.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results