Paper Contents
Abstract
This project explores an adaptive rule-based machine translation system designed for translating English sentences into Telugu. The proposed approach utilizes a combination of rule-based methodologies, including if-then logic for optimal rule selection, probability-based word choice, and rough set theory for sentence classification. The system relies on a set of production rules, a comprehensive training set, and a bilingual dictionary for both English and Telugu.The translation process begins with tokenizing the input English sentence into individual words, which are then tagged with their respective parts of speech (POS). Words not present in the predefined database are tagged using formulated grammatical rules. By leveraging these POS tags, the system retrieves appropriate word translations from the database and concatenates them to form the final translated sentence in Telugu.The motivation for developing this translation system stems from several key factors: the scarcity of robust translation systems from English to Indian languages and the specific linguistic complexities of Telugu, which features intricate phrasal, word, and sentence structures. Additionally, while direct machine translation (MT) is often used for related languages, this work applies it to the more challenging Telugu-to-English translation, aiming for simplicity, rapid development, and enhanced accuracy.
Copyright
Copyright © 2024 B SANJANA. This is an open access article distributed under the Creative Commons Attribution License.
