The Dynamic Polymer Annotated Library: An automated approach to curating materials science literature

It remains challenging to curate data about the design, properties, and applications for a given class of materials across thousands of papers to identify emerging trends and efficiently train new artificial intelligence (AI) models. Here, we present a universal strategy for the automated curation and analysis of an annotated library of over 10 5 identified papers focused on a specific class of materials. Our streamlined two-step review process begins with a traditional Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) scoping review to create a database of relevant papers. This is followed by an automated tagging procedure that annotates papers based on key molecular design features, properties, and applications. We demonstrate this approach on the emerging field of dynamic polymers by generating the Dynamic Polymer Annotated Library (DPAL) and analyzing it through example case studies. Our approach enables extraction of design-relevant information about the material structure, properties, and applications.