Year: 2014
Type: conference-paper
Abstract: This article is a system description paper and reports on the participation of Techlimed in the "QALB-2014 shared task" on evaluation of automatic arabic error correction systems organized in conjunction with the EMNLP 2014 Workshop on Arabic Natural Language Processing.Correcting automatically texts in Arabic is a challenging task due to the complexity and rich morphology of the Arabic language and the lack of appropriate resources, (e.g.publicly available corpora and tools).To develop our systems, we considered several approaches from rule based systems to statistical methods.Our results on the development set show that the statistical system outperforms the lexicon driven approach with a precision of 71%, a recall of 50% and a F-measure of 59%.
DOI: 10.3115/v1/w14-3621
Authors: Djamel MOSTEFA, Omar ASBAYOU, Ramzi ABBES
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W2251943172