Reranking with Linguistic and Semantic Features for Arabic Optical Character Recognition
Nadi Tomeh, Nizar Y. Habash, Ryan M. Roth, Noura Farra, Pradeep Dasigi, Mona Diab · 2013
Optical Character Recognition (OCR) sys-tems for Arabic rely on information con-tained in the scanned images to recognize sequences of characters and on language models to emphasize fluency. In this paper we incorporate linguistically and seman-tically motivated features to an existing OCR system. To do so we follow an n-best list reranking approach that exploits recent advances in learning to rank techniques. We achieve 10.1 % and 11.4 % reduction in recognition word error rate (WER) relative to a standard baseline system on typewrit-ten and handwritten Arabic respectively. 1