Replication package for "Protocol Reverse Engineering Using Large Language Models: A Systematic Mapping Study"

This dataset is the replication package for the systematic mapping study "Protocol Reverse Engineering Using Large Language Models: A Systematic Mapping Study". It contains the data needed to trace the search, study selection, supplementary citation searching, data extraction and figures reported in the paper. Contents: - 01_search: search strings for IEEE Xplore, ACM Digital Library, Scopus and Web of Science; search log; query refinement log. - 02_screening: title/abstract screening decisions of both reviewers for all 68 unique records; inter-rater agreement (Cohen's kappa = 0.77); summary of exclusion reasons; full-text screening of the 9 retained reports. - 03_supplementary_search: PRISMA counts; backward and forward snowballing records; Scopus and Web of Science cited-by checks. - 04_extraction: data extraction for the 9 included studies (corpus frozen on 14 September 2026) and the final coding as reported in the paper. - 05_figures: PRISMA flow diagram; mapping figure (PDF) and the Python script that generates it. Notes: - The database searches were run on 11 September 2026. - The Scopus search string was reconstructed from the archived search log; it uses the same terms as the Web of Science string, searched in TITLE-ABS-KEY. - Exclusion reasons were recorded in aggregate, not per record. - Full texts of the included studies, abstracts and raw database exports are not included for licensing reasons. - The package was assembled with the assistance of Anthropic Claude; all contents were checked by the authors.