[Corpora-List] New corpus announcement: UKPConvArg1 - convincingness of Web arguments

Habernal, Ivan habernal at ukp.informatik.tu-darmstadt.de
Mon Jun 13 14:37:19 CEST 2016

Dear corpora users,

We are publicly releasing a new data set for assessing convincingness of Web arguments written in natural language, the UKPConvArg1 corpus. Evaluating qualitative aspects of argumentation represents a new challenge in Computational Argumentation, an emerging area of NLP.

UKPConvArg1 corpus consists of 11,650 argument pairs - two arguments with the same standpoint to the given topic, manually annotated with a binary relation describing which argument from the pair is more convincing. Each pair also contains several reasons written in natural language explaining which properties of the arguments influence their convincingness.

We introduce the UKPConvArg1 corpus in detail in our upcoming ACL 2016 paper:

Habernal, I. & Gurevych, I. (2016). Which argument is more convincing? Analyzing and predicting convincingness of Web arguments using bidirectional LSTM. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (ACL 2016), Volume 1: Long Papers. Pages: to appear. Berlin, Germany. Association for Computational Linguistics.


The corpus is licensed under a permissive CC-BY license and can be downloaded here:


Related experimental software is also being released at our GitHub page:


Feel free to contact me, should you have any questions.

Best regards,

Ivan Habernal

-- ----------------------------------------------------------------------------- Ivan Habernal, Ph.D. Postdoctoral Researcher Ubiquitous Knowledge Processing (UKP) Lab FB 20 / Computer Science Department Technische Universitšt Darmstadt Hochschulstr. 10, D-64289 Darmstadt, Germany Room S2/02/B107 habernal at ukp.informatik.tu-darmstadt.de www.ukp.tu-darmstadt.de

Web Research at TU Darmstadt (WeRC) www.werc.tu-darmstadt.de

Adaptive Preparation of Information from Heterogeneous Sources (AIPHES) GRK 1994 / www.aiphes.tu-darmstadt.de

Knowledge Discovery in Scientific Literature (KDSL) PhD program / www.kdsl.tu-darmstadt.de -----------------------------------------------------------------------------

More information about the Corpora mailing list