WebNLG Challenge 2017 Results

Data | Challenge Results | Research | Corpus Documentation

The WebNLG Challenge 2017 was held in April-October 2017. Below you will find the content as it was posted in 2017.

Challenge Summary

Participant Data

Download .zip with all the submissions, teams' reports, and a baseline output; or download the same data per team.

Automatic Evaluation Results

  • The WebNLG Challenge: Generating Text from RDF Data. C. Gardent, A. Shimorina, S. Narayan, L. Perez-Beltrachini. Proceedings of INLG 2017. PDF
  • Scripts to reproduce results are here.
  • Together with the results, we release a gold standard only for unseen (new) DBPedia categories, i.e. the half of the gold standard. Download here.
  • All test data with references (seen and unseen categories together) is now released. Download here.
  • NEW: A Jupyter notebook with automatic results can be found here. (credit: Abelardo Vieira Mota)

Human Evaluation Results

  • The WebNLG Challenge: Report on Human Evaluation. A. Shimorina, C. Gardent, S. Narayan, L. Perez-Beltrachini. PDF-v1 PDF-v2 (corrected version)
  • Human scores, references and scripts are here.

Final Release (Larger Dataset)