People around the world speak different languages and so involves of the translation of one language into another. These are now done by means of a process called machine translation which uses some specific software to complete the tasks. This is done by the substitution of words in initial stages which may then lead to the translation of the entire language. This can be done more efficiently through the various statistical techniques which depend upon the collection and organization of data. This technique has proved convenient and also improves the quality of data by doing several statistical models. The translation process involves two main steps which are defined as the decoding and encoding process. The decoding process finds the meaning of the given text and on the other side the encoding process converts the particular text into the target language. Linguistic rules are being used for the efficient conversion and are based completely on the knowledge of the human beings. There are many types of translations which are categorized as rule based and dictionary based translations. Each of these technologies will first analyses the given text and converts it into a proper syntax for further translation process.