Study-level Evidence Extraction and Forensic Nano–AI Maturity Assessment (FNAI) Dataset for a Systematic Evidence Map of Nano–AI in Forensic Science

Description

This dataset supports the systematic evidence map entitled “Nano–AI Convergence in Forensic Science: A Systematic Evidence Map of Sensing, Analytics, Validation, and Forensic Decision Support.” The review evaluates the convergence of nanotechnology-enabled forensic sensing and artificial intelligence/machine learning. The dataset contains original bibliographic search exports, study-level evidence extraction, eligibility and screening audit information, the Forensic Nano–AI Maturity Assessment (FNAI) coding framework and matrix, and aggregate synthesis tables. Searches of Scopus, PubMed/MEDLINE, IEEE Xplore, and ScienceDirect yielded 217 records, which were reduced to 163 unique records and ultimately 24 CORE-NANO–AI studies after screening and full-text assessment. The dataset is intended to support transparency, reproducibility, and independent inspection of the evidence synthesis.

Steps to reproduce

The dataset was generated through a systematic evidence-mapping workflow following PRISMA 2020 and PRISMA-S principles. Searches were conducted in Scopus, PubMed/MEDLINE, IEEE Xplore, and ScienceDirect for records published from 1 January 2005 through 7 September 2026. The raw authenticated database exports are preserved in this dataset. Records were combined and deduplicated using normalized DOI matching and, where DOI was unavailable, normalized title matching, followed by screening and full-text eligibility assessment. The final evidence set comprised 24 CORE-NANO–AI studies. Study-level evidence was extracted using predefined fields covering nanotechnology platform, forensic evidence type, AI/ML method, datasets, validation, performance, robustness, reproducibility, portability, explainability/uncertainty, and forensic interpretability/readiness. The Forensic Nano–AI Maturity Assessment (FNAI) framework was then applied using predefined ordinal scores from 0–3 across 11 dimensions. Aggregate synthesis tables were generated from the coded study-level dataset. Raw third-party database exports should be interpreted subject to the terms and conditions of their respective database providers.

Institutions

• Aditya University

  Andhra Pradesh, Sūrampālem

Categories

Licence