<![CDATA[The Bima language is one of the regional languages in Indonesia that still lacks support in natural language processing technology. The application of Transformer architecture for regional languages has not been widely researched, especially for the Bima language. This study aims to develop an automatic translation system from Bima to Indonesian using a Transformer-Based Neural Machine Translation (NMT) approach. The methods used include the collection and processing of parallel corpora, training NMT models using the OpenNMT framework, and evaluating translation results using metrics such as BLEU and TER. The data used will be collected from various sources, including manually translated texts by linguists and available local documents. The model evaluation results indicate that the model was successfully developed and can translate sentences well. This approach is expected to become the foundation for the development of automatic translation technology for other regional languages in Indonesia, while also contributing to the preservation and digitization of local languages through artificial intelligence technology.]]>
Copyrights © 2025
