Subproject 4.4: The Knowledge Map of the Early Modern Period

Project Manager: Prof. Dr. Wilhelm Kühlmann
Department of German Studies, University of Heidelberg

Since 1996 the projects MATEO, CAMENA and TERMINI have digitized selected books of the early modern period (1450-1750) and offer scans and machine readable texts on the Internet together with a number of tools for analyzing them. About 80 percent of the texts are written in Latin of the early modern period, which has not undergone any systematic linguistic analysis yet. The TERMINI project collects keywords that are treated in reference works. It also extracts related data and stores them along with the lemma concerned. A meta dictionary (LEMMATA) has been set up, which combines semantic, morphological and syntactic information from different lexical and encyclopedic sources. On this basis, co-occurrence analysis as used on the context of modern text mining approaches automatically extracts typical contexts of a lemma. These contexts are interpreted as indicating the meaning of a lemma. Ambiguities too can be identified in cases where those contexts split up in two or more semantically different sets. The results can be presented in form of a list of contexts for each lemma. Additionally, they can be visualized as graphs, giving an intuitive picture of the semantic space of an entire text or corpus in form of a knowledge map. Thus providing a set of typical contexts of a lemma empirically checking and evaluating standard semantical assumptions is possible as well. Finally, comparing co-occurrence data from the early modern corpus with a corpus of ancient texts makes changes visible caused by certain epistemical developements in the early modern period.