Create comperative_sentence

#1

If I am not mistaken, the dataset does only contains different sentences aligned as AI ( 0 ) or Human ( 1 ). I think for pragmatic detection mechanism, we could develop a dataset that the same sematics to be written in terms of different pragmatics by AI and Humans and use those labeled datas for Semi-Supervised learning

Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment