“…Apart from the five mentioned extraction tasks, it can be easily extended to tasks related to extracting information from scientific literature and transforming data into a simple user-friendly reaction format 22 that is both human- and machine-readable. This approach will significantly contribute to the development of extensive databases like the Open Reaction Database, 23,24 SciFinder 25 and Reaxys, 26 which gather comprehensive synthesis data through automated curation and expert verification, to make data more findable, accessible, interoperable, and reusable (FAIR).…”