Key points are not available for this paper at this time.
The Philippines boasts a rich array of languages, each with its unique cultural significance and history.Despite their importance as a vital component of the country's linguistic heritage, the growth and development of these low-resource languages has been impeded by the lack of parallel corpusa valuable tool in machine translation.This article presents a comprehensive process of constructing a Bikol-Filipino parallel corpus, commencing from web scraping to sentence and word alignment.The study underlines the significance of evaluating and validating the corpus to ensure its accuracy and reliability.The collected data was utilized to fine-tune a T5-Base transformer model for machine translation and subsequently assessed with the BLEU metric.The resulting score of 73.71 highlights the significance of the generated Bikol-Filipino parallel corpora, making it an invaluable asset for research and development in both languages.
A 2024 study studied this question.