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
Lately I am developing NN based systems for Dependency Parsing and POS Tagging task for Turkish using UD Treebank (v 2.9).
I wanted to include your two data files in my training sets, however there are many differences between your labels (both POS and Dependency labels) and UD labels.
Would you consider offering a utility function to convert them to UD standards? (Although I guess some are not subset of UD and conversion may not be possible? e.g: mwe)
Nevertheless, thank you for the hand tagged gold dataset.
The text was updated successfully, but these errors were encountered:
Thank you for using our dataset. That's correct, our segmentation and the tagset is currently not UD compatible. We do have a plan to create a UD version of this dataset, but due to low bandwidth it is not clear when we'll get to it.
As you suggest, it is likely that a fully automated conversion is not going to be possible due to some subtle differences.
Lately I am developing NN based systems for Dependency Parsing and POS Tagging task for Turkish using UD Treebank (v 2.9).
I wanted to include your two data files in my training sets, however there are many differences between your labels (both POS and Dependency labels) and UD labels.
Would you consider offering a utility function to convert them to UD standards? (Although I guess some are not subset of UD and conversion may not be possible? e.g: mwe)
Nevertheless, thank you for the hand tagged gold dataset.
The text was updated successfully, but these errors were encountered: