pls cf. sect. 5 : I think it is not suited to get coocurrent B3Kat concepts, which are identifiers for the records in B3Kat library catalogue. The goal is to compute the coocurrence of a GND to other GNDs and Subject-Identifiers from other Subject-Vocabularies (RVK, DDC, Lccn). Hence, I propose to compute the coocurrence of concepts directly and attach it to the GND-Graphs from Step 1. This will result in many triples but we may cut of adding coocurrece relation triples to a GND-Graph (Hanns Eisler) at a certain degree of coocurrence, i.e. 5-10%...
pls cf. sect. 5 : I think it is not suited to get coocurrent B3Kat concepts, which are identifiers for the records in B3Kat library catalogue. The goal is to compute the coocurrence of a GND to other GNDs and Subject-Identifiers from other Subject-Vocabularies (RVK, DDC, Lccn). Hence, I propose to compute the coocurrence of concepts directly and attach it to the GND-Graphs from Step 1. This will result in many triples but we may cut of adding coocurrece relation triples to a GND-Graph (Hanns Eisler) at a certain degree of coocurrence, i.e. 5-10%...