K.H. van Dalen-Oskam
J. de Does
M.J. Marx
I. Sijaranamual
K. Depuydt
B. Verheij
V. Geirnaert
Date (dd-mm-yyyy)
Named Entity Recognition and Resolution for Literary Studies
Computational Linguistics in the Netherlands Journal
Publication Year
Document type
Faculty of Science (FNWI)
Informatics Institute (IVI)
This paper reports on the project Namescape: Mapping the Landscape of Names in Modern Dutch Literature, funded by CLARIN-NL. The background of the project is research in literary onomastics, the study of the usage and functions of proper names in literary (i.e. fictional) texts. The two main tasks for the project were to adapt existing Named Entity Recognition software to modern Dutch
fiction, and to perform Named Entity Resolution by linking the names to
Wikipedia entries. For Named Entity Recognition, existing tools have been
trained on literary texts and a new NE tagger has been developed. The standard
list of name categories had to be extended, since the analysis of the usage of
proper names in literature needs to distinguish e.g. between first names and
family names. The Named Entity Resolution task was done to explore the
possibility of labeling the names in fiction in another way, by categorizing a
name as referring to a person or location that only exist in the story of a
fictional work (plot-internal names), or one referring to a person or location
in the real world (plot-external names). This distinction is linked to the
hypothesis that plot-internal and plot-external names can have different
(stylistic and narrative) functions. Automatically marking them up is the first
step toward testing that hypothesis on a large corpus. In this paper we describe
the results of these two main tasks.