We aim to generate the data necessary for polymer informatics by developing information extraction pipelines to automatically extract polymer properties from the literature. We use natural language processing software (ChemDataExtractor) to extract both names of chemical entities and properties automatically from journal articles.
Efforts are a collaboration between NIST and CHiMaD, specifically computer scientists and polymer physicists at Univ. of Chicago. We are also currently looking at ways of hosting data directly provided by others.
Ongoing Challenges