A Conceptual Framework for Human-in-the-Loop Multi-attribute Decision Making Under Uncertainty
FLINS-ISKE 2026, Sydney, Avustralya, 15 - 19 Temmuz 2026, cilt.16756, ss.205-223, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Cilt numarası: 16756
- Doi Numarası: 10.1007/978-981-92-2497-5_12
- Basıldığı Şehir: Sydney
- Basıldığı Ülke: Avustralya
- Sayfa Sayıları: ss.205-223
- Galatasaray Üniversitesi Adresli: Evet
Özet
Decision-making in complex environments is often characterized by multiple decision criteria, uncertainty, and the interaction between humans and intelligent systems. In recent years, human-in-the-loop (HITL) artificial intelligence has emerged as an effective approach to improve the transparency, trustworthiness, and adaptability of decision processes by leveraging the synergy between humans and intelligent agents or machine learning models. At the same time, multi-criteria decision-making (MCDM) has become a widely used paradigm for addressing problems involving multiple, often conflicting criteria. Despite growing interest in both fields, their integration in uncertainty-aware environments remains relatively underexplored in the literature. This paper presents a structured survey of recent advancements at the intersection of HITL AI and MCDM under uncertainty. It reviews fundamental concepts in human-centric AI, examines recent studies combining human expertise with intelligent decision-support systems, and highlights application domains such as healthcare, finance, and smart infrastructure. Key challenges—including interpretability, trust, bias mitigation, and human–AI collaboration—are also discussed. A conceptual framework is outlined to illustrate how HITL mechanisms can be integrated into uncertainty-aware MCDM processes. The contribution of this work is primarily conceptual and integrative, aiming to structure existing knowledge and outline future research directions for the development of human-centered intelligent decision-support systems.