WING-NUS/scisumm-corpus

data corruption in J96-3004.xml

Closed this issue · 2 comments

jungokasai commented

Hi, I just found that one example in Training Set 2017 is not properly assigned sids for the abstract.

cmkumar87 commented

@jungokasai I confirm that the abstract of the said training file does not have sids assigned for the Abstract section. Thanks for the report!

However, it is inconsequential since none of the annotated reference spans are from the Abstract section. In other words, all the annotated references spans here: https://github.com/WING-NUS/scisumm-corpus/blob/master/data/Training-Set-2017/J96-3004/annotation/J96-3004.ann.txt have an sid assigned.
Please let us know if you find it otherwise.

  • 👍1
jungokasai commented

Right, thank you for your quick response!

  • 👍1