Data

We used the pre-processed datasets (ONB and LFT) from the A Named Entity Recognition Shootout for German paper. Unfortunately, the datasets are not publicly available. Thus, you have to contact Martin Riedl to obtain the dataset.

Once you've obtained the datasets, put all files in this folder here. We expect the following files:

• enp_DE.lft.mr.tok.{train,dev,test}.bio - LFT dataset
• enp_DE.onb.mr.tok.{train,dev,test}.bio - ONB dataset