VEA Model in Word Formation Process of Maithili MT (original) (raw)

Morphological analysis is the most remarkable stage for the development of Maithili-English-Hindi MT system under NLP. This paper is motivated to design a morph analyzer for Maithili language and as add-on of Maithili MT system for appropriate analysis at the morphological level. The research is contributing through derivational process of analyzing word attached with affixes. This paper has reviewed most of the methods of MA at different level. Among all the development of MA, suffix striping and FSA are commonly practiced. Some of them are using lemma based approach for analyzing morph (Nikhil et.al. 2012). There are several linguists and computer scientists have discussed the statistical approach and Hybrid approach with FSA, probability based model (Rinju et. al. 2013) and several other approaches of machine learning to develop Maithili Morphological Analyzer (MMA). But rule based approach is friendly enough to use with all the machine learning models. VEA is one of the emerging modal in word formation process which is somehow adequate with Maithili language. Overall linguistics approach is core of MA in Maithili for developing MT system. This research is introducing linguistically a friendly and bit new with machine learning and proficient model for analyzing words and generating multiple words on the basis. The discussion also covered the concatenation with root word to suffix and prefix. Maithili MA is demonstrating a small concept with rule based model and we are designing it with hybrid modal including corpus based approach. This design is incorporating the lexicon tables, suffix list, prefix list and the Vowel Ending Approach (VEA) to justify that how does concatenation take place. In the above table POS category of Noun, Adjective and Verb are shifting to another category of POS after concatenation of suffixes. And it's also focused that how the words end with their vowels and how does suffix connects on the basis of its vowel ending mechanism.