• For the use of this code you will likely want to have a manual parse or a parser (e.g. [udpipe][http:https://ufal.mff.cuni.cz/udpipe]).
• To extract m2 files, one would like to already have m2 files (possibly with non-sense types), those can be extracted manually or automatically (e.g. [errant][https://github.com/chrisjbryant/errant])

See Using_m2, currently the file creates both matrices and m2 in the main function, comment out if unneeded (matrix is get_confusion_matrix + extract_matrices) python using_m2/GEC_UD_divergences_m2.py source.conllu reference.conllu m2file.m2

See Using_m2, currently the file creates both matrices and m2 in the main function, comment out if unneeded (m2 is in syntactic_m2 function) python using_m2/GEC_UD_divergences_m2.py source.conllu reference.conllu m2file.m2

There is also an end to end implementation for English with or without combining with ERRANT (adaptations to support other languages or morphology when POS is not changing are welcome)

python using_m2 preprocessing.py m2file.m2

If you have found this work useful please cite the CoNLL paper.

    title = "Classifying Syntactic Errors in Learner Language",
    author = "Choshen, Leshem  and
      Nikolaev, Dmitry  and
      Berzak, Yevgeni  and
      Abend, Omri",
    booktitle = "Proceedings of the 24th Conference on Computational Natural Language Learning",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/2020.conll-1.7",
    pages = "97--107",
    }