Alkali-Activated Slag Database

The repository includes a cleaned literature-derived AAS database containing 3,053 compressive strength records collected from 136 independent studies, covering 155 chemically distinct ground granulated blast-furnace slag (GGBS) sources. Each record corresponds to one AAS mixture at a specified curing age with associated precursor chemistry, mixture proportions, curing and testing conditions, specimen geometry, and measured compressive strength. Each record contains the following groups of attributes: GGBS chemical composition and physical properties, including SiO₂, Al₂O₃, CaO, MgO, Fe₂O₃, Na₂O, K₂O, MnO, Mn₂O₃, TiO₂, loss on ignition, Blaine fineness, and the calculated GGBS reactivity descriptor AMODE (average metal oxide dissociation energy); Mixture proportions, including Na₂O/GGBS, silicate modulus, superplasticizer/GGBS, water/GGBS, fine aggregate/GGBS, and coarse aggregate/GGBS; Curing conditions, including base curing temperature, additional curing temperature and time, relative humidity, logarithm-transformed curing energy descriptor, curing age, and logarithm-transformed age; Specimen geometry, including aspect ratio, specimen volume, and equivalent length; Measured compressive strength; Source information, including the original source DOI when available. The database was curated through manual source verification and model-assisted screening. Duplicate records were removed, and flagged observations were checked against the original sources and the database’s inclusion criteria. Selected missing values were imputed or inferred from reported information, as described in the manuscript. The database includes source references and DOIs where available to support traceability and further research. Details of the database construction, feature selection, reactivity calculation, and data cleaning procedures are reported in the associated publication, “Reactivity-Informed Machine Learning for Performance Prediction and Design Space Exploration of Alkali-Activated Slag”.