Oriya morphological analyzer using lttoolbox

Abstract

A Morphological analyzer is an essential tool for many NLP applications. Developing a fully-fledged morphological analyzer (MA) tool for an agglutinative language like Oriya is a challenging task. This paper deals with development of a MA for Oriya, a resource poor language. The MA is being developed using the paradigm approach. It consists of various paradigms under which nouns, pronouns, adjectives, verbs and indeclinables are classified. The paradigms have been created for inflected forms using an XML based morphological dictionary from the Lttoolbox package. At present, a total of 10,840 words have been entered into the dictionary. In the course of the paper we talk about the design and implementation of the MA. We also talk about the issues and limitations experienced in developing it using Lttoolbox.