Computational linguistic tools and machine translation system for Kannada language

Abstract

Developing a complete and well specified natural language processing (NLP) models for any language with limited electronic resources is always a challenging and demanding task. The informationssuch as news, weather report, annual report, technical and scientific books etc. are currently available only in a selected number of languages especially in English in the Internet.But people who know only the local language may not be able to use these resources, though the resources are available. This signifies the need for providing information in regional languages to the user. There were genuine reasons and motivations behind each and every good and recognizable work. The proposed work was motivated by the following factors. South Dravidian language like Kannada is having almost 40 million speakers and has its own independent script and long documented histories. Even though Kannada is a language rich in literature, its resources are poor when viewed through the prism of computational linguistics. The developmentofNLP in Kannada language is not explored much and is in the beginning stage compared to other Indian languages. Moreover, Kannada is a highly agglutinative and morphologically rich language. Syntactic and semantic variance makes the problem much harder for creating full fledged computational linguistic tools and machine translation (MT) system for Kannada language. With the absence of computational linguistic tools and English to Kannada MT system so far, even a reasonable domain specific development can find its immediate applications in government and educational sectors. The central goal of proposed work was to develop a Rule Based Machine Translation (RBMT) systemfor English to Kannada language by integrating different modules and various computational linguistic tools.