EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Hadis Araştırmalarında Büyük Dil Modelleri (LLM): Teorik Çerçeve, Sınırlar ve Yapay Zekâ Destekli Yapılandırılmış Veri İnşası

Orhan Çinici

Mevzu – Sosyal Bilimler Dergisi · 15.06.2026 · (16) · 299-326

DOI: 10.56720/msbd.2026.s16.326 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Mevzu – Sosyal Bilimler Dergisi, online erschienen am 15.06.2026. Verfasst von Orhan Çinici. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This article examines whether large language models (LLMs) can serve as a verifiable tool in the field of hadith studies. The study focuses on three fundamental structural issues encountered by general-purpose models in domain-specific applications: corpus dependency, the risk of hallucination, and the limitations of the context window. To address these issues, the article proposes a solution framework that integrates constraining system prompts, a Retrieval-Augmented Generation (RAG) architecture, and a vector database tailored to the field. In the theoretical section, conducted using a qualitative descriptive method, the relevant literature is reviewed. In the applied section, Volume 23 of al-Dhahabī’s Siyar Aʿlām al-Nubalāʾ was transferred to a vector database and converted into a structured dataset using the Gemini 2.5 Flash model. The study produced a rijāl dataset comprising 21 variables for 271 scholars. The findings indicate that, when constrained by restrictive guidelines, the model minimized the risk of hallucination: rather than generating fabricated information, it clearly reported missing data and successfully analyzed the complex structure of Arabic rijāl literature. Accordingly, AI systems supported by data-driven architectures and supervised by subject-matter experts appear to offer significant opportunities as a complementary tool in comprehensive hadith research, including rijāl and isnād analysis, textual analysis, the extraction of narrator-relationship networks, and the identification of ṭarīq and ṭabaqa.

Zitieren

APA

Çinici, O. (2026). Hadis Araştırmalarında Büyük Dil Modelleri (LLM): Teorik Çerçeve, Sınırlar ve Yapay Zekâ Destekli Yapılandırılmış Veri İnşası. Mevzu – Sosyal Bilimler Dergisi, 299-326. https://doi.org/10.56720/msbd.2026.s16.326

Vancouver

Çinici O. Hadis Araştırmalarında Büyük Dil Modelleri (LLM): Teorik Çerçeve, Sınırlar ve Yapay Zekâ Destekli Yapılandırılmış Veri İnşası. Mevzu – Sosyal Bilimler Dergisi. 2026:299-326. doi:10.56720/msbd.2026.s16.326

BibTeX
@article{inici2026,
  author = {Çinici, Orhan},
  title = {Hadis Araştırmalarında Büyük Dil Modelleri (LLM): Teorik Çerçeve, Sınırlar ve Yapay Zekâ Destekli Yapılandırılmış Veri İnşası},
  journal = {Mevzu – Sosyal Bilimler Dergisi},
  year = {2026},
  number = {16},
  pages = {299--326},
  doi = {10.56720/msbd.2026.s16.326}
}
Kostenlos starten →