The role of Artificial Intelligence (AI) in revitalizing low-resource languages (The Kurdish language as an example)

Life Cycle Reliability and Safety Engineering · Springer Nature · 2026-10-01 · DOI: 10.1007/s10772-026-10295-4

Аннотация

This research uses the Kurdish language to critically evaluate AI and NLP technologies for revitalization and preserving speech in low resource environments. Following the PRISMA methodology, this research analyzes existing resources and research on the Kurdish digital speech and text corpora. It reviews the state of the art (SotA) of AI technologies and their subsystems, such as KuBERT, Conformer-based ASR, and VITS-based TTS synthesis. It identifies current gaps in available resources for major Kurdish dialects (Central, Northern, Southern, and Gorani/Hawrami). This research also develops a strategic planning framework for digitally empowering Kurdish and incorporating it into global initiatives. This planning framework supports the local open-source development of ecosystems and preservation of different dialects and literacy levels with cross cutting approaches of transfer learning. Finally, this research links Kurdish digital innovations with concepts of public discourse and ethical considerations. The study draws upon examples of ‘algorithmic bias’ and ‘data sovereignty’ as guiding principles for the construction of ethical AI systems. This research combines the development of various Kurdish AI systems to help overcome inadequate resources while preserving rich diversity of the dialects.

Читать у издателя →

Похожие статьи