Rapid Adaptation of NE Resolvers for Humanities Domains using Active Annotation

Abstract

The entities mentioned in collections of scholarly articles in the Humanities (and in other scholarly domains) belong to different types from those familiar from news corpora, hence new resources need to be annotated to create supervised taggers for tasks such as ne extraction. However, in such domains there is a great need for making the best use possible of the annotators. One technique designed for this purpose is active annotation. We discuss our use of active annotation for annotating corpora of articles about Archaeology in the Portale della Ricerca Umanistica Trentina.