The Deutscher Novellenschatz as a use case in topic modeling <

conducted by: Thomas Wetin


The corpus of the Deutscher Novellenschatz contains 86 texts of 82 male and female authors enriched with metadata. The quantitative analysis is taking into account the historical problem to which the collection is a reaction, that is, the outcomes of literary mass production. The question of the possibility of distinction in the context of mass similarity is relevant both in terms of genre poetics as well as for the experimental methods of literary history production, which lies in the foundation of this novel collection. Topic modeling is aimed at helping to sort the findings from stylometric and network analysis of global similarities and local similar groups in the corpus.