Araştırma Makalesi

A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems

Cilt: 14 Sayı: 2 30 Haziran 2026
PDF İndir
EN TR

A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems

Öz

Integrating AI into OIS poses several challenges in terms of governance due to the possibility of algorithmic bias. In ERP systems, AI-driven modules play a critical role in making key decisions related to employee evaluation, resource management, demand prediction, and procurement of suppliers. With biased data, erroneous modeling assumptions, and black-box decision-making, the impact could be adverse to particular user categories, skewing organizational processes and potentially harming the company's reputation and regulatory compliance. Even though there is increased consciousness about AI bias, no systematic criteria have been put forward for prioritizing bias factors in OIS. This paper proposes a novel hybrid approach to identifying and prioritizing AI bias criteria in OIS through a combination of literature review and Fuzzy-AHP. The expert opinions of five AI and OIS specialists were used to generate criterion weights under fuzziness conditions. The findings reveal three main bias factors: Machine Learning (w = 0.5304), Fairness and Ethics (w = 0.3043), and Natural Language Processing (w = 0.1653). From the list of sub-criteria, Training Data Quality (GW = 0.2430), Model Explainability (GW = 0.1345), and Accountability and Transparency (GW = 0.1304) are the most significant factors. The suggested methodology provides a framework for aiding OIS design, IT management, and organizational policy formulation on AI bias governance.

Anahtar Kelimeler

Kaynakça

  1. [1] Ahmed, F., & Kilic, K. (2019). Fuzzy Analytic Hierarchy Process: A performance analysis of various algorithms. Fuzzy Sets and Systems, 362, 110-128.
  2. [2] Atan, M., Atan, S., & Altın, K. (2008). İnsan Kaynakları Seçiminde Analitik Hiyerarşi Süreci Kullanımı Ve Bir Yazılım Önerisi. Gazi Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 10(3), 143-162..
  3. [3] Arsu, T., & Varlı, E. (2022). Çok kriterli karar verme yöntemiyle finansal performans analizi: İtfaiye malzemeleri ve çok maksatlı yangın müdahale-kurtarma araçları ihracatı yapan bir firma uygulaması. Düzce Üniversitesi Sosyal Bilimler Dergisi, 12(2), 305-325.
  4. [4] Barocas, S., Hardt, M., & Narayanan, A. (2023). Fairness and machine learning: Limitations and opportunities. MIT Press.
  5. [5] Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of stochastic parrots: Can language models be too big? Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency (pp. 610-623).
  6. [6] Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J. D., Dhariwal, P., ... & Amodei, D. (2020). Language models are few-shot learners. Advances in Neural Information Processing Systems, 33, 1877-1901.
  7. [7] Buolamwini, J., & Gebru, T. (2018). Gender shades: Intersectional accuracy disparities in commercial gender classification. Conference on Fairness, Accountability and Transparency (pp. 77-91).
  8. [8] Calabrese, A., Costa, R., Levialdi, N., & Menichini, T. (2016). A fuzzy analytic hierarchy process method to support materiality assessment in sustainability reporting. Journal of Cleaner Production, 121, 248-264. https://doi.org/10.1016/j.jclepro.2015.12.005

Ayrıntılar

Birincil Dil

İngilizce

Konular

Bilgi Sistemleri Organizasyonu ve Yönetimi

Bölüm

Araştırma Makalesi

Erken Görünüm Tarihi

19 Haziran 2026

Yayımlanma Tarihi

30 Haziran 2026

Gönderilme Tarihi

7 Nisan 2026

Kabul Tarihi

4 Mayıs 2026

Yayımlandığı Sayı

Yıl 2026 Cilt: 14 Sayı: 2

Kaynak Göster

APA
Asiloğulları Ayan, M. (2026). A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems. Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji, 14(2), 903-913. https://doi.org/10.29109/gujsc.1924721
AMA
1.Asiloğulları Ayan M. A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems. GUJS Part C. 2026;14(2):903-913. doi:10.29109/gujsc.1924721
Chicago
Asiloğulları Ayan, Merve. 2026. “A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems”. Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji 14 (2): 903-13. https://doi.org/10.29109/gujsc.1924721.
EndNote
Asiloğulları Ayan M (01 Haziran 2026) A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems. Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji 14 2 903–913.
IEEE
[1]M. Asiloğulları Ayan, “A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems”, GUJS Part C, c. 14, sy 2, ss. 903–913, Haz. 2026, doi: 10.29109/gujsc.1924721.
ISNAD
Asiloğulları Ayan, Merve. “A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems”. Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji 14/2 (01 Haziran 2026): 903-913. https://doi.org/10.29109/gujsc.1924721.
JAMA
1.Asiloğulları Ayan M. A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems. GUJS Part C. 2026;14:903–913.
MLA
Asiloğulları Ayan, Merve. “A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems”. Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji, c. 14, sy 2, Haziran 2026, ss. 903-1, doi:10.29109/gujsc.1924721.
Vancouver
1.Merve Asiloğulları Ayan. A Hybrid Framework for Identifying and Weighting Artificial Intelligence Bias Criteria in Organizational Information Systems. GUJS Part C. 01 Haziran 2026;14(2):903-1. doi:10.29109/gujsc.1924721

                                     16168      16167     16166     21432        logo.png   


    e-ISSN:2147-9526