On the Optical Character Recognition and Machine Translation Technology in Arabic: Problems and Solutions

The report addresses the basic problems of the Arabic language formalization based on analysis of linguistic errors in software products. Reviewing the principles of modern information systems operation the authors come to the conclusion that the existing methods of the Arabic formalization allow to note a shift towards the technological aspects of the linguistic processing of facts, however, the quality of applied linguistic components still remains poor. Possibilities for the application of traditional recognition algorithms for Arabic are still uncertain in spite of a significant number of theoretical and practical results in the field of computational linguistics. There are several problems which are due to be solved in relation to the processing of the Arabic text. These issues may be divided into those related to Optical Recognition of the Written Text (OCR), Word Processing (WP) and building of the content of the dictionaries, Machine Translation (MT).