Doing More with Named Entities

Abstract

The usability, disclosure and value of digitized full text collections can be improved by linking named entities or events in text to linked data [1]. These data can be used to obtain additional information and it can be used in queries for expressing conditions in terms of semantic relations. When the links are obtained automatically, we need disambiguation by users or by sophisticated algorithms. Providers of full text data will have to deal with all of these aspects. In this paper we discuss our plans and work in progress and propose a common approach to increase interoperability between text data from different providers. We feel that it should be the ambition of every provider of text collections to have as many named entities as possible in text identified with globally unique persistent identifiers, linked to one or more resource descriptions to guide users through the enormous amount of digitized text. Notice: This work reflects the personal view of the authors and does not yet reflect current policies of the KB. © Springer International Publishing Switzerland 2014.