Speech Recognition based on Saudi Arabia Dialect

Modern Standard Arabic . First , colloquial Arabic is divided into Western and Eastern Arabic . The Western Arabic is further subdivided into Libyan , Tunisian , Moroccan , and Algerian Arabic while the Eastern Arabic is subdivided into Gulf , Egyptian , and Levantine dialects . The former is composed of Jordanian , Syrian , Iraqi , and Lebanese sub groups . However , this categorization does not automatically imply that the Arabic dialects are distinct in their specific placements . As a matter of fact , one subgroup can still showcase different dialects depending on whether the people reside in the urban or rural regions

or the kind of tribe in which they belong (Sakarna , 2005 : 524 . These groups determine ways in which a dialect is formed such that in an Arabic dialectical subgroup like Lebanese , it is found out that people from the north , south , Beirut and mountainous regions of Lebanon differ in their dialects despite all of them being under the subgroup of Lebanese

This can also be categorized as a language morphology problem as the Arabic language has a continual growth rate of vocabulary making it difficult t come up with standard language models for the Arabic dialects . Sakarna (2005 : 525 ) assert that competition within the various dialects and the need for a particular group to appear the most sophisticated is a contributing factor to the development of various dialects

Hyassat and Zitar (2006 : 136 ) summarize the major challenges with Arabic language speech recognition . First , the Arabic dialect lacks short vowels and other information to pronunciation in most Arabic texts Second , the morphological complexity of the Arabic language makes it difficult to establish an ideal speech recognition system . Third , the discrepancies between the formal and the dialectical Arabic also pose a major challenge to research concerned about speech recognition systems .Motivation for Arabic ASR systems and advantages of the speech recognition...