Development of morpholoical analyzer for Kannada verbs

Morphological analyzer is a tool which identifies and analyzes the internal structure of given word and also gives the morphological and grammatical information associated with the given word. Kannada is an inflectional, derivational, agglutinative and morphologically very rich language. Developing a well-fledged morphological analyzer (MA) for Kannada is a challenging task. In this paper, a hybrid MA for Kannada verbs is proposed as a part of machine translation system from Kannada to any other language. This model is developed using suffix-stripping, rule-based and paradigm-based approaches. The performance of this model is tested against a set of verbs randomly taken from a well-known dictionary called “Kannada Rathna Kosha” [1].