We propose a semantic linking model for the Turkic language family, most members of which are considered low-resource from the computational linguistics viewpoint. The theoretical foundation of the model is Liddy’s layered scheme of natural language processing, compressed here into four levels: graphemic-morphological, syntactic, semantic, and cross-lingual. Rather than treating expert-built knowledge resources as hard constraints, we introduce them as a noisy supervision signal weighted by a per-language reliability coefficient ????????; this construction subsumes both classical retrofitting (???????? → 1) and purely distributional models (???????? → 0) as special cases of a single scheme. The optimization problem is augmented with a fairness constraint bounding the gap between the best and worst per-language quality, which prevents the model from drifting toward resource-rich languages at the expense of low-resource ones. A two-level algorithm is derived that combines Adam updates for embeddings, a closed-form orthogonal Procrustes step for cross-lingual mappings, and softmax reparametrisation for the language-weight simplex with a projected subgradient step for the dual variable; a standard ????(1/√????) bound is stated for the outer loop. The evaluation protocol includes five baselines that isolate the contribution of each model component, intrinsic metrics (Spearman correlation on SimRelUz, Precision@1 and Precision@5 for bilingual lexicon induction) and extrinsic metrics (chrF++ for downstream machine translation), with 95% bootstrap confidence intervals and Holm–Bonferroni correction for multiple comparisons.
| Mualliflar | Akhmedjanova, D.A., Ахмеджанова, Д.А. |
|---|---|
| Jurnal | Ҳисоблаш ва амалий математика муаммолари |
| Nashriyot | Тошкент ахборот технологиялари университети қошидаги, Ахборот-коммуникация технологиялари илмий-инновацион, маркази |
| ISSN | 2181-8460, 2181-046X |
| Nashr sanasi | 2026-09-15 |
| Son | 4 |
| Betlar | 171-181 |
| Til | Ingliz |
| DOI | 10.71310/pcam.4_74.2026.11 |
| Indeksga qo‘shilgan | 2026-09-16 |
DOI: 10.71310/pcam.4_74.2026.11 · Maqolaning asl sahifasi
Turkic languages, cross-lingual embeddings, low-resource languages, noisy supervision, fairness constraint, Procrustes problem, тюркские языки, кросс-язычные векторные представления, малоресурсные языки, зашумлённый контроль, ограничение справедливости, задача Прокруста
| APA | Akhmedjanova, D.A., & Ахмеджанова, Д.А. (2026). A semantic linking model for Turkic languages: a fairness-constrained cross-lingual approach for low-resource settings. Ҳисоблаш ва амалий математика муаммолари, (4), 171-181. https://doi.org/10.71310/pcam.4_74.2026.11 |
|---|---|
| GOST (OAK) | Akhmedjanova, D.A., Ахмеджанова, Д.А. A semantic linking model for Turkic languages: a fairness-constrained cross-lingual approach for low-resource settings // Ҳисоблаш ва амалий математика муаммолари. – 2026. – № 4. – P. 171-181. – DOI: 10.71310/pcam.4_74.2026.11. |
| MLA | Akhmedjanova, D.A., and Ахмеджанова, Д.А. “A semantic linking model for Turkic languages: a fairness-constrained cross-lingual approach for low-resource settings.” Ҳисоблаш ва амалий математика муаммолари, (4), 171-181, 2026. https://doi.org/10.71310/pcam.4_74.2026.11 |
| Chicago | Akhmedjanova, D.A., and Ахмеджанова, Д.А. “A semantic linking model for Turkic languages: a fairness-constrained cross-lingual approach for low-resource settings.” Ҳисоблаш ва амалий математика муаммолари, (4), 171-181 (2026). https://doi.org/10.71310/pcam.4_74.2026.11 |
| Harvard | Akhmedjanova, D.A., and Ахмеджанова, Д.А (2026) ‘A semantic linking model for Turkic languages: a fairness-constrained cross-lingual approach for low-resource settings’, Ҳисоблаш ва амалий математика муаммолари, (4), 171-181. https://doi.org/10.71310/pcam.4_74.2026.11 |
| BibTeX | @article{akhmedjanovada2026,
title = {A semantic linking model for Turkic languages: a fairness-constrained cross-lingual approach for low-resource settings},
author = {Akhmedjanova, D.A. and Ахмеджанова, Д.А.},
journal = {Ҳисоблаш ва амалий математика муаммолари},
year = {2026},
number = {4},
pages = {171-181},
doi = {10.71310/pcam.4_74.2026.11},
url = {https://doi.org/10.71310/pcam.4_74.2026.11}
} |
Boshqa maqola uchun iqtibos tayyorlash
This article presents a mathematical model of suspension filtration in a filter section, taking into account the saturation of each layer of the porous medium. An incompressible fluid in free-flowing regions is…
Classical quadrature rules do not always account simultaneously for the weight, node geometry, and functional class of the integrand, which motivates formulas optimized with respect to the norm of the error functional…
Numerical integration in Sobolev spaces often relies on optimal quadrature formulas in the sense of Sard to minimize the calculation error. While many standard formulas exist, constructing optimal rules that incorporate…
We study a family of degenerate-singular second- and third-order elliptic boundary value problems modeling anomalous diffusion in heterogeneous tumor tissue, all built on one differential operator with a power-law…
This paper presents a complete mathematical model for the automatic extraction of phraseological units (idioms) from texts in Turkic languages, with a focus on Uzbek. The task is formulated as a binary classification…
Ҳисоблаш ва амалий математика муаммолари — barcha maqolalar · 2026-yil arxivi
This paper formalizes and extends the classic concept of the "Middle Turkic"language (Karimov–Mutalov, 1992) by applying tools from corpus and computational linguistics. The proposed model frames the selection of an…