Argument Graphs Extracted From the Online Debate Portal Kialo

For our paper "Towards an Argument Mining Pipeline Transforming Texts to Argument Graphs" published at COMMA 2020, we extracted debates from the online debate portal Kialo. We used it to train multiple classifiers and evaluate our pipeline, but the data may be useful in other areas as well. The argument graphs are saved in a format similar to the one used by OVA (https://ova.arg-tech.org). They can be parsed using our Python library (https://pypi.org/project/recap-argument-graph/).

If you use the data, please cite our work as follows:

@inproceedings{Lenz2020ArgumentMiningPipeline,
address = {{Perugia, Italy}},
archiveprefix = {arXiv},
author = {Lenz, Mirko and Sahitaj, Premtim and Kallenberg, Sean and Coors, Christopher and Dumani, Lorik and Schenkel, Ralf and Bergmann, Ralph},
booktitle = {Proceedings of the 8th {{International Conference}} on {{Computational Models}} of {{Argument}}},
doi = {10.3233/FAIA200510},
editor = {Prakken, Henry and Bistarelli, Stefano and Santini, Francesco and Taticchi, Carlo},
eprint = {2006.04562},
eprinttype = {arxiv},
pages = {263--270},
publisher = {{IOS Press}},
series = {Frontiers in {{Artificial Intelligence}} and {{Applications}}},
title = {Towards an {{Argument Mining Pipeline Transforming Texts}} to {{Argument Graphs}}},
volume = {326},
year = {2020},
Bdsk-Url-1 = {https://doi.org/10.3233/FAIA200510}}