Annotated textual dataset PV600 of perovskite bandgaps for information extraction from literature

The data generation details and computational framework for this dataset can be found in the manuscript Sipilä, M., Mehryary, F., Pyysalo, S., Ginter, F. & Todorović, M. (2025) Annotated textual dataset PV600 of perovskite bandgaps for information extraction from literature. This dataset contains 600 annotated text snippets about five perovskite materials bandgaps. The dataset is intended to be used in evaluating the information extraction methods for materials science. The text snippets were extracted from open access articles downloaded using query word “perovskite" from five publishing sources: Elsevier, Springer, Royal Society of Chemistry, Core API and arXiV. The dataset comprises five materials, MAPI, FAPI, MAPB, CsPbI3 and CsPbBr3. In total, there are 227 annotated bandgap values derived from 188 snippets (see article above for details). The annotated dataset is made available in two formats: the csv-table format and in the brat annotation format.