Title: DESIGN AND IMPLEMENTATION OF INTERACTIVE WEB SYSTEM FOR THE KAZAKH TEXT RECOGNITION AND CORRECTION WITH USING OF PARALLEL COMPUTING

Issue Number: Vol. 3, No. 4
Year of Publication: 2013
Page Numbers: 100-114
Authors: Bakyt M. Kairakbay, Daniyar B. Nurseitov, Yuriy Y. Stolyarov, Ilyas E. Tursunov, Igor A. Ugay, David L. Zaurbekov
Journal Name: International Journal of New Computer Architectures and their Applications (IJNCAA)
- Hong Kong

Abstract:


In given paper we present the integrated interactive computer recognition system for the Kazakh language text with using of parallel computing. The design and integration methodology of the system are based on service-oriented architecture that allows provide an easy, flexible, and extensible integration of any language service into any desktop or mobile client. 4-tier SOA was designed and built based on W3C Web service standard. Using the high-performance cluster demonstrates significant advantages in the Kazakh text processing, especially for large arrays of texts. The main objective of the developing system is to provide any person easy access to text and documentation in Kazakh language with the following possibility of editing and manipulation different docs through respective Web services OCR and morphological analysis/correction. Developed web services cover the Kazakh language text OCR and morphological analysis for subsequent correction of errors after OCR, and show an acceptable quality of the Kazakh texts recognition that is better than existing tools of text recognition.