Scalable and Explainable AI Architectures for Intelligent Decision-Support Systems: Systematic Literature Review Protocol

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Protocol of a systematic literature review, registered before the searches are run. The review maps how AI-based decision-support systems integrate data pipelines, machine-learning and relational inference, explainability, decision-support interaction, and human oversight, to ground a reference architecture for scalable and explainable decision support. It follows Kitchenham and Charters (2007) and will be reported according to PRISMA 2020. The protocol defines three research questions (layers integrated; how explainability and human oversight are realized; levels of evaluation), six databases (Scopus, Web of Science, IEEE Xplore, ACM Digital Library, arXiv, OpenAIRE Graph) plus Google Scholar and citation searching, five supervisor-approved search strings with five documented supplementary strings and their pilot results, eligibility criteria, a double-screened 20% sample with Cohen's kappa, a five-item quality assessment, a data-extraction form, and the synthesis plan. Appendix A gives the exact query for every database.