Vom Thesaurus zum semantischen Netz. Potenziale von Data Mining in bibliographischen Datensätzen

Abstract

The paper explores the potential of data mining in bibliographic records. Quantitative analsysis of co-occurrences of keywords in bibliographic records makes it possible to create a semantic network. In contrast to the rigid hierarchy of traditional thesauri, this network maps polyvalent semantic relations between keywords across the distinct thematic branches of a thesaurus tree. Analysis of a comprehensive data pool made up from two major bibliographies (Gnomon Bibliographic Database and ZENON DAI) should allow for the semantic mapping of a core set of central topics (people, places, social structures, concepts) of classical scholarship.