An argument-annotated corpus of scientific publications: Resource analysis

General Description

This repository provides statistics on the newly added argumentation annotations (components + relations) and on the relationship between all the label sets available in the argument-extended Dr. Inventor Corpus.

The Dr. Inventor Corpus was originally compiled by [1].

Requires python-igraph for the argumentation graph analysis: http://igraph.org/python/.

Obtaining the data

The corpus is available here: http://data.dws.informatik.uni-mannheim.de/sci-arg/compiled_corpus.zip

Giving Credit

If you use the corpus or the code in any way please cite our corresponding publication:

A. Lauscher, G. Glavaš und S. P. Ponzetto, “An Argument-Annotated Corpus of Scientific Publications”, in Proceedings of the 5th Workshop on Argument Mining at EMNLP 2018, Brussels, Belgium: Association for Computational Linguistics, 2018, 40‒46.

BibTeX:

@inproceedings{lauscher2018,
  address = {Brussels, Belgium},
  title = {An Argument-Annotated Corpus of Scientific Publications},
  booktitle = {Proceedings of the 5th {{Workshop}} on {{Argument Mining}} at {{EMNLP}} 2018},
  publisher = {{Association for Computational Linguistics}},
  author = {Lauscher, Anne and Glavaš, Goran and Ponzetto, Simone Paolo},
  year = {2018},
  url = {https://madoc.bib.uni-mannheim.de/46084/},
  pages = {40--46}
}

References

[1] Beatriz Fisas, Francesco Ronzano, and Horacio Saggion. 2016. A multi-layered annotated corpus of scientific papers. In Proceedings of the International Conference on Language Resources and Evaluation, pages 3081–3088, Portoroz, Slovenia. European Language Resources Association.