Developing Parallel Corpus for the Machine Translation System in Dzongkha Language

The primary objective of this work is to create a parallel corpus for a machine translation system for the Dzongkha language. The ability to communicate across cultures and languages is made possible through machine translation, which is essential for breaking down language barriers. The availability of high-quality parallel corpora, which are bilingual resources made up of aligned sentences in the source and target languages, is what determines how effective the method is. Due to its limited resources and unique linguistic features, Dzongkha, the native language of Bhutan, presents a difficult task for machine translation. The creation of trustworthy machine translation systems for Dzongkha is severely hampered by the lack of a complete parallel corpus. Therefore, the purpose of this work is to close this gap by creating a parallel corpus that is customized exclusively for Dzongkha. As part of the study approach, parallel sentences from Dzongkha and one or more source languages are gathered and aligned. We will use a variety of data collecting methods, such as Internet crawling, Optical Character Recognition (OCR), reusing already completed translations, and hiring native speakers to do the translations. To build a bilingual corpus appropriate for training machine translation models, the phrases will be manually aligned. Linguistic specialists and native speakers will be involved in the validation and verification process to guarantee the corpus's correctness and quality. The numerous areas and genres of the Dzongkha language will also be carefully captured in order to increase the system's adaptability. The parallel corpus that has been created will be an invaluable tool for training Dzongkha-specific machine translation models. The findings of this study will help enhance machine translation technology for Dzongkha, allowing cross-linguistic interaction, encouraging cultural exchange, and assisting different Bhutanese fields.