metadata
license: unknown
task_categories:
- text-classification
language:
- en
size_categories:
- 10K<n<100K
A dataset of contradictory sentence pairs.
This is the dataset from a Master's thesis: https://repositorio-aberto.up.pt/bitstream/10216/129029/2/415679.pdf. The name of the dataset comes from the title of the thesis. You can find some documentation for the dataset in the following repository: https://github.com/BeatrizBaldaia/sentence-pair-contradictions/tree/master
Originally collated by Beatriz Souto de Sá Baldaia, I take no credit for that aspect, I just processed and uploaded it here because I needed to train a sentence contradiction model.
This dataset is a merge of several others, so I'm not sure what the license should be.
I think there is quite a lot of political content in the data, but models trained from it do seem to generalise