In this paper we present aMarathipart of speech tagger. Itismorphologically rich language. itis spokenby the native peopleofMaharashtra. The general approachused for development of tagger isstatisticalusingTrigram Method.Themain concept ofTrigramis toexplore the most likely POS for a token based ongiven information of previous two tags bycalculatingprobabilitiesto determine whichthe best sequence oftag is. Inthis paper we show the development of the tagger. Moreoverwe have alsoshown the evaluationdone
展开▼