Constructing a Machine Translation (MT) system is a multidimensional activity which demands innovation and problem-solving ability in several areas. In this process, certain inter-disciplinary fields of science and technology are vastly utilized. Strategies of Artificial Intelligence (AI), through application principles of mathematics, statistics and logic provide an appropriate form of language which can be programmed and transferred into the machine. In Iran, the aforementioned activity has been initiated of late but bears an inconsistent progress. In order to achieve advancements in the field of MT, three tasks should be taken into account: 1) data-processing, describing and constructing suitable models for intelligent applications, 2) writing and applying the required software, and 3) utilizing the relevant technology. Taking these into consideration, in this article, some linguistic aspects of the MT system are explicated by placing premium on methods experimented in Russia. Moreover, due to the integrated relationship of MT with AI and programming, some general information, with regard to the named fields, is also provided.
)