BRIN Develops AI-Based Multilingual Translation Model To Preserve Bugis Language
JAKARTA The National Research and Innovation Agency (BRIN) developed a prototype of the Artificial Intelligence (AI)-based translation model in a new language. The institution will add Bugis.
By developing a multilingual translation for Bugis, BRIN wants to preserve endangered regional languages. The initiative that utilizes the development of this technology is believed to prevent the loss of cultural heritage.
Indonesia has about 718 regional languages, but its use is decreasing because people tend to use Indonesian or foreign languages. The less regional language speakers, the language is threatened with extinction.
Seeing this, BRIN through the Data and Information Science Research Center (PRSDI) developed an AI-based regional language model. They want to preserve linguistic wealth by utilizing the Big Language Model (LLM).
SEE ALSO:
In developing this AI-based multilingual translation, BRIN collaborated with the Faculty of Cultural Sciences, Hasanuddin University and the Faculty of Computer Sciences, Handayani University. This time, the focus of the three is a prototype for Bugis.
By developing multilingual translations, people can convert Bugis to foreign languages such as English, Spanish and Russian, and vice versa. This translation works in text to text, audio, and video formats to text.
The platform developed by BRIN currently has implemented around 14,000 Bugis words. To view the translated words, the public can access the prototype through the webhttp://nlp-prototype.brin.go.id/ website.
The number of words added in Bugis may increase in the future. However, this additional vocabulary will not occur in a short time because BRIN is facing data limitations and has difficulty finding the right regional linguist.