An experience in developing the Nepali sense tagged corpus

A key resource that aids in several NLP tasks is WordNet. Wordnet is used as the sense inventory for sense tagging of corpus. Sense tagging is the task of tagging each word in the sentence with the correct sense of the word in the given context. Sense tagging activity helps in validation of WordNet and improvement of Wordnet quality. Sense tagging is one of the toughest annotation works and this paper discusses about the Sense Tagging tool, procedures involved in sense tagging the Nepali corpus and the challenges involved in sense tagging. Nepali WordNet is used as the sense inventory for sense tagging of Nepali corpus. For accurately sense tagging voluminous data, a standard and definitive lexicon is required. In this work the corpus in Nepali language is taken from newspaper domain.