Labelled corpus and qualitative-subset extraction tables for A Survey of EEG Applications in Transport: A Taxonomy and Critical Analysis

This deposit contains the four data files that underpin a 2026 survey of 261 EEG-in-transport publications (2020–2026). Each paper is classified along three orthogonal axes — application domain (Driver, BCI, Operator, Pedestrian/Crowd, Other), methodology (Deep Learning, Classical Machine Learning, Signal Processing, Hybrid, Other), and dataset type (Simulator, Real-world, Public benchmark, Not specified). A stratified subset of 90 papers was read in full and characterised against a 19-field extraction template. Files included: EEG_Transport_Labelled.csv — full 261-paper corpus with bibliographic metadata, abstract, tags, and three-axis labels assigned from titles, abstracts, and tags. EEG_Transport_Labelled_v2.csv — the same 261-paper corpus with an additional "Dataset Type (v2)" column that reports the dataset-type label assigned after the full-text re-screening described in Section 2.5 of the manuscript. The original "Dataset Type" column is preserved for auditability; 92 of 143 abstract-level "not specified" papers received an updated label. EEG_Transport_Qualitative90.csv — the 90-paper qualitative subset with selection ranks, in-stratum rarity scores, and the diversity-score breakdown described in Section 2.4 (Equation 1). EEG_Transport_Detailed90.csv — the 19-field detailed extraction populated from full-text reads of the 90 qualitative-subset papers, covering specific architecture, feature-extraction approach, EEG channel count and acquisition device, sample size, dataset, peak accuracy and other metrics, cross-validation strategy, author-stated limitations, and a one-sentence contribution summary. All numerical claims, figures, and tables in the manuscript can be reproduced from these CSVs. The associated manuscript is currently under review at Transportation Research Part F: Traffic Psychology and Behaviour.