Towards building standard datasets for Arabic recognition | |
---|---|
( Volume 2 Issue 2,February 2016 ) OPEN ACCESS | |
Author(s): | |
Mohamed E. M. Musa | |
Abstract: | |
Current Machine learning and pattern recognition method needs big dataset to produce efficient and accurate recognizers. The lack of standard big Arabic dataset is one of the big challenges that faces the research in this area. This paper presents the Arabic dataset collected and annotated by SUST-ALT (Sudan University of Science and Technology- Arabic Language Technology group) to contribute in filling this gap. The datasets contain: numerals datasets, isolated Arabic letters datasets, Arabic names datasets. Tese datasets contain offline dataset as well as online ones. The paper also describes some published results as well as future work. |
|
Paper Statistics: | Total View : 506 | Downloads : 497 | Page No: 16-19 | |
Cite this Article: | Click here to get all Styles of Citation using DOI of the article. |