The phases or the stages through which the process of machine translation gets completed is:
The very first phase in the machine translation process is the text input. The sentence categories are classified based on the degree of difficulty of translation. This is because of the inter-relationship among adjacent sentences.
This phase makes the machine translation process easier and qualitative. The source language text may consist of figures, flowcharts, etc that do not require any translation. Hence, only those portions that need translation are identified. Once the text gets translated, the target text needs to be reformatted after post-editing. Reformatting is performed to check if the non-translation process is still there in the text or not.
The level of pre-editing and post-editing depends on the efficiency of the specific Machine Translation system. For some systems, it may be required to segment the long sentences into short sentences. Fixing the punctuation marks and blocking the matter that does not require translation are also done during pre-editing. Post editing is required to make sure that the quality of the translation is fine enough
The analysis phase determines the word form such as inflections, tense, number, part of speech, etc. Syntactic and semantic analysis are mostly executed simultaneously to produce syntactic tree structure and semantic network respectively. The sentence generation phase is simply the reverse of the process of analysis.
Computational morphology technique deals with analysis, recognition, and generation of words. Some of the morphological processes are inflection, derivation, affixes, and combining forms. Inflection is considered the most regular and productive morphological process across all the languages. The role of inflection is to alter the form of the word in number, gender, mood, tense, aspect, person, and case. The morphological analyzer gives information concerning morphological properties of the words it analyses.
Syntactic Analysis is concerned with how the words get grouped into classes, which are called parts-of-speech, and the way in which words depend on other words in a sentence.
Tagging refers to the identification of linguistic properties of the individual words and parsing refers to the assessment of the functions of the words in relation to each other.
A semantic analysis job is to compose the meaning representations and assign them linguistic inputs. The source of knowledge contains the meaning of words, meanings associated with grammatical structures, and knowledge about the discourse context.