Analysis of family relationships of characters in the srpELTeC corpus based on semantic web techniques and textometry

Abstract

This paper presents an analysis of family relationships of characters in the srpELTeC collection of Serbian novels published between 1840 and 1920, using a textometric approach and semantic web techniques. The study explores the appearance of family relationships between literary characters in the corpus in relation to the periods in which the novels were written, as well as the influence of the author. Considering the overlap between the literary movements of Romanticism and Realism during this period, the analysis examines which direction family relationships were more prevalent in and whether male characters are more represented than female characters in the collection. Visualizations of the relationships between characters are presented through linking the characters with novels in Wikidata and by issuing SPARQL queries, as well as through network analysis using the Gephi tool.