You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
build_features.py line 111:
Is this line trying to replace the original corpus? I think it does nothing now. This information is also used in function: preprocess_corpus_notDropEntity
can I replace that line with this? corpus[int(index2ind[int(ind)])] = corpus[int(index2ind[int(ind)])].replace(ori, ent)
The text was updated successfully, but these errors were encountered:
Hi there, sorry for late reply. We tried changing that line. But the difference is not much (1% change in f1 score). We are not getting good score for our dataset. Graph provided by the author works fine.
I have a simple question. For agnews dataset, do we use the title or the description for the "document"?
build_features.py line 111:
Is this line trying to replace the original corpus? I think it does nothing now. This information is also used in function:
preprocess_corpus_notDropEntity
can I replace that line with this?
corpus[int(index2ind[int(ind)])] = corpus[int(index2ind[int(ind)])].replace(ori, ent)
The text was updated successfully, but these errors were encountered: