Extraction dataset, search strings, and PRISMA flow data for "Constrained multi-agent reinforcement learning for distributed resource management: a review"

This dataset contains the complete, auditable record behind a systematic literature review (SLR) following the PRISMA 2020 protocol combined with the Kitchenham guidelines. It synthesises 24 peer-reviewed journal articles (2021–2026) on Constrained Multi-Agent Reinforcement Learning (CMARL) for distributed resource management, and lets any reader reproduce the study selection and verify every per-article claim made in the paper. Files: (1) the full data-extraction workbook (24 articles × 57 fields, plus the Scopus and IEEE PRISMA screening records); (2) the verified Boolean search strings for Scopus and IEEE Xplore, including three supplementary check searches added during journal revision; (3) the full PRISMA identification/screening/eligibility/inclusion counts with per-criterion exclusion breakdown; (4) a completed PRISMA 2020 27-item reporting checklist; (5) raw Scopus export data (CSV) for the three supplementary check searches — quantifying retrieval bias from search-term choice and the effect of excluding conference papers — each independently reproducible from the query strings in file (2); (6) a README describing the dataset structure and classification axes.