A simple script to preprocess raw and annotation files of TED corpus and save to tsv
Assuming the folders are created correctly, convert_ted_to_csv.py creates flat tsv version of TED-MBD. It has only limited fields -
- rel_type
- rel_sense
- conn_text
- arg1_text
- arg2_text