AI SAFEGUARDING OF AZERBAIJANI IN DIGITAL SPACES
Shusha Scientific Journal · 26.02.2026 · 2 (1) · 17
DOI: 10.66366/ssj.2026.40 ↗Kurz gefasst
Fachartikel in der Zeitschrift Shusha Scientific Journal, online erschienen am 26.02.2026. Verfasst von Narmina Aliyeva und 3 weiteren Autor:innen.
Abstract (Originalsprache)
The rapid expansion of digital communication platforms has intensified multilingual interaction and accelerated non‑standard language practices, posing distinctive challenges for languages with comparatively limited digital resources. This study examines Azerbaijani as a representative case for understanding how low‑resource languages can be safeguarded in digital spaces without suppressing naturally occurring variation. Integrating perspectives from digital sociolinguistics, philology, and computational linguistics, the research investigates patterns of code‑switching and orthographic variation in Azerbaijani online discourse and evaluates their implications for normativity and language‑technology development. Empirically, the study analyzes publicly available Azerbaijani social‑media texts alongside a curated gold‑standard reference subset representing standard written usage. Code‑switching is examined through hybrid language‑identification techniques, while orthographic variation is classified using a phonologically informed taxonomy. These analyses are operationalized within a reproducible benchmarking pipeline that combines orthographic normalization, probabilistic language identification, and norm‑sensitive evaluation metrics. The results demonstrate that code‑switching and non‑standard spellings in digital Azerbaijani are systematic, socially motivated, and structurally constrained rather than arbitrary or degradative. Orthographic variation frequently reflects phonological reduction and efficiency‑oriented writing strategies, enabling high normalization recovery rates when appropriately modeled. Benchmarking outcomes further show that linguistic variability and computational robustness are compatible when evaluation frameworks account for digital registers and gradient normativity. By linking empirical findings with contemporary language‑governance discourse, the study argues that safeguarding a language in digital environments is best achieved through evidence‑based documentation, transparent benchmarks, and supportive language technologies rather than prescriptive restriction. Beyond Azerbaijani, the proposed framework offers a transferable model for low‑resource languages seeking sustainable digital visibility and technological inclusion in multilingual online ecosystems.
Zitieren
APA
Aliyeva, N., Khudiyeva, G., Ismayil, Z., & Asadova, M. (2026). AI SAFEGUARDING OF AZERBAIJANI IN DIGITAL SPACES. Shusha Scientific Journal, 2(1), 17. https://doi.org/10.66366/ssj.2026.40
Vancouver
Aliyeva N, Khudiyeva G, Ismayil Z, Asadova M. AI SAFEGUARDING OF AZERBAIJANI IN DIGITAL SPACES. Shusha Scientific Journal. 2026;2(1):17. doi:10.66366/ssj.2026.40
BibTeX
@article{Aliyeva2026,
author = {Aliyeva, Narmina and Khudiyeva, Gunash and Ismayil, Zulfiyya and Asadova, Meliha},
title = {AI SAFEGUARDING OF AZERBAIJANI IN DIGITAL SPACES},
journal = {Shusha Scientific Journal},
year = {2026},
volume = {2},
number = {1},
pages = {17},
doi = {10.66366/ssj.2026.40}
}
EUVIMED