Development of a rule-based Yorùbá numerals to digits translator

The importance of numbers in commerce, education, and everyday conversation cannot be overemphasised. With increasingcross-border information exchange, the need to accurately translate numbers between languages of the world becomesessential. While significant research has focused on the translation of Hindu-Arabic numerals into Yorùbá text, theconversion of Yorùbá numeral text to digits has not been explored by the research community, and the translation qualitiesprovided by existing general-purpose machine translation models and Large Language Models that support Yorùbálanguage remain very poor. This paper presents a rule-based approach for the translation of Yorùbá numerals into digits. Ituses a hierarchical replacement algorithm for token and operator substitution, and the Reverse Polish Notation (RPN)evaluation method with stack-based parsing to process complex Yorùbá numeral expressions. The system achieved 100%accuracy on a comprehensive test suite of over 100,000 Yorùbá numeral expressions curated from different sources thatgenerated the numbers through rule-based system. In addition, the system achieved an accuracy of 97.64% on noisy dataextracted from publicly accessible Yorùbá corpora, demonstrating the viability of rule-based approaches for this task. Thiswork addresses a critical gap in Yorùbá language technology, providing a foundational component for text-to-speechnormalisation, voice-enabled applications, and bidirectional numeral translation systems. The model is made available asan open-source resource to support further research in Yorùbá computational linguistics