Abstract
There are millions of people in the world speak many languages. To communicate with each other it is necessary to know the language which we use. To do this operation we use language identification system.
In general, Automatic Speech Recognition for English and other languages has been the subject of most researches in the last forty years. Arabic language research has been growing very slowly in comparison to English language research. The Arabic language has many different dialects; they must be identified before Automatic Speech Recognition can take place.
This paper describes the design and implementation of a new spoken language identification system: Arabic Spoken Language Identification (ASLIS). It focuses only on two major dialects: Modern Standard Arabic (MSA) and Egyptian. It presents a spoken Arabic identifier using Hidden Markov Models (HMMs), and it is developed using the portable Hidden Markov Model Toolkit (HTK).