What's in an entity? Exploring Nested Named Entity Recognition in the Historical Land Register of Basel (1400-1700).

In this presentation I present my work in which I explore the application of nested named entity recognition (NNER) techniques on a specific corpus, that of the Historical Land Records of Basel, a repository of property transactions, dating from late medieval to early modern times. I evaluate an established NNER system and compare it to a self-developed approach utilizing the Flair NLP framework, showing that even on comparatively small amount of training data, reliable annotation can be achieved with my self-developed system. These results lay the groundwork for further experiments in the realm of information extraction, such as extraction relations and events from pre-modern documents. 

The links to the presented model and code can be found in the file "repositories.md".