This research investigates sound length in speech, comparing its physical and mental representation across languages. While English treats length as phonetic variation, languages like Japanese use it meaningfully. The study focuses on Persian to improve speech recognition and therapy, helping determine how sound duration is perceived, produced, and processed cognitively.

This research addresses the exclusion of minority and low-resource languages from modern language technologies. Using linked data principles, it builds interconnected, machine-readable linguistic resources for languages like Cree, Welsh, and Kurdish. The goal is to enable inclusive AI systems and future technologies that support global communication across diverse linguistic communities.