October 07, 2013
LitSOM – Automatic Mapping of Russian Authors using SOM

Master thesis in Applied Computer Science at U of Salzburg


The LitSOM proposed in this Master thesis applies Self-Organizing Maps (SOM) and Learning Vector Quantization (LVQ) to Russian literature. The SOM is used to create a map of contemporary Russian novels, which enables literary scholars to study intertextual relationships. Thus, LitSOM can be regarded as an example for the emerging scientific discipline of digital humanities. The quality of the resulting maps of Russian literature is assessed subjectively, i.e., from a philologist’s point of view, and objectively, i.e., in a „classic“ problem of text mining, namely, author classification. For that purpose the SOM can be easily converted to implement LVQ.

Download LitSOM presentation.