Entities database
This json file contains the named entities extracted from 3.27 million materials science abstracts.
Each document is indexed by it's digital object identifier (DOI) which may be used to find the original article. Each document contains the following entity types: material (MAT), sample descriptor (DSC), symmetry/phase label (SPL), property (PRO), application (APL), synthesis method (SMT), and characterization method (CMT).
The named entity recognition was optimized for inorganic materials science.
