A Dataset for Arabic Textual Entailment
Maytham Alabbas · Research Explorer (The University of Manchester) · 2013
There are fewer resources for textual entailment(TE) for Arabic than for other languages,and the manpower for constructingsuch a resource is hard to come by.We describe here a semi-automatic techniquefor creating a first dataset for TEsystems for Arabic using an extension ofthe ‘headline-lead paragraph’ technique.We also sketch the difficulties inherent involunteer annotators-based judgment, anddescribe a regime to ameliorate some ofthese.