Towards an Automated Approach to Extract and Compare Fictional Networks: An Initial Evaluation

Within the field of Digital Humanities, fictional networks are defined by the social interactions among characters, capturing unique features associated with each different textual source. The ability to automatically extract, analyse, and assess such properties would open new research directions, and lead to relevant and important results. However, these are challenging tasks which require a multi-disciplinary approach due to their inherent complexity. In this paper, we will describe the initial implementation of a system to address the above. This is part of an ongoing research investigation aiming to define and implement a similarity measure to compare specific properties associated with fictional networks extracted from across different genres and sources, as well as from different historical periods. The preliminary evaluation clearly shows the potential of our approach, as well as the applications to a variety of multi-disciplinary research fields.