The IDs you stripped from the dependencies in remove_id() should stay there. If two identical words occur in the same sentence, and you strip the word-id from the results, there's no way for us to easily disambiguate them (hence, why Stanford explicitly put them there)
The IDs you stripped from the dependencies in remove_id() should stay there. If two identical words occur in the same sentence, and you strip the word-id from the results, there's no way for us to easily disambiguate them (hence, why Stanford explicitly put them there)