It would be very beneficial, for potential extension of the gamma, to have the corpus shuffling tool introduced in this paper:
Manual Corpus Annotation: Giving Meaning to the Evaluation Metricslink
There used to be available code link
It allows to test for different type of errors:
False negative
False positive
Fragmentation (Which I think the Pygamma is not taking into account and would be very beneficial to account for)
It would be very beneficial, for potential extension of the gamma, to have the corpus shuffling tool introduced in this paper:
Manual Corpus Annotation: Giving Meaning to the Evaluation Metrics
link There used to be available code link It allows to test for different type of errors: