Use different tokenizer for claims to wikipedia articles

sheffieldnlp / naacl2018-fever

Fact Extraction and VERification baseline published in NAACL2018

http://fever.ai

Apache License 2.0

127 stars 41 forks source link

Closed j6mes closed 6 years ago

j6mes commented 6 years ago

Wikipedia articles are pre-tokenized and just need splitting by space. Claims need to be tokenized properly