Test set split

[McKracken]

Hi, is there a way to obtain the test set annotation example IDs (or the name of the 23 documents) you used for your experiments on FrameNet 1.5, especially related to the paper "Semi-Supervised Frame-Semantic Parsing for Unknown Predicates", where you used 2,420 sentences from the fulltext annotations?