Ein regelbasierter Ansatz für eine Type-Token-Zuordnung in historischen Wörterbuchbelegen

Abstract Wortgeschichte digital (‘Digital Word History’) is a historical dictionary of the German language that focuses on describing semantic shifts from about 1600 through today. Hallmarks of the project are the narrative form of its articles and the fact that most of them deal with more than one lemma at a time. Due to the narrative form, the quotations are separated from the text as not to disturb the line of argument. The online presentation of the articles does not only provide text and quotations but tries to enhance them with a variety of technical tools among which is the Belegnavigator (‘quotation navigator’). The navigator visualizes the distribution of the quotations and the first occurrence of the article lemmas on a timeline. In doing so, the automatic visualization makes use of canonicalization strategies in order to properly conflate the lemma tokens in the quotations with their types.