A Dataset for Arabic Textual Entailment

Abstract

There are fewer resources for textual entailment (TE) for Arabic than for other languages, and the manpower for constructing such a resource is hard to come by. We describe here a semi-automatic technique for creating a first dataset for TE systems for Arabic using an extension of the ‘headline-lead paragraph’ technique. We also sketch the difficulties inherent in volunteer annotators-based judgment, and describe a regime to ameliorate some of these.

4 Figures and Tables

Cite this paper

@inproceedings{Alabbas2013ADF, title={A Dataset for Arabic Textual Entailment}, author={Maytham Alabbas}, booktitle={RANLP}, year={2013} }