Ryan abstract

This paper describes some of the work of the ‘Networking Archives’ project, which has put together a dataset of almost half a million letter records from the Tudor and Stuart State Papers as well as the resource ‘Early Modern Letters Online'. It describes and evaluates the process of using transfer learning to retrain a Named Entity Recognition (NER) model capable of extracting place names from the short abstracts of the State Papers, and how the resulting placenames were then geocoded with a custom method using contextual information. Finally, the paper demonstrates how the project used this data—and a web application which can be used to map and browse the archive through the places mentioned—to understand more about one particular section of the State Papers: the extensive network set up by the Undersecretary of State, Joseph Williamson, which provided him with handwritten news.