Towards a model for replicating aesthetic literary appreciation

Abstract

This study aims to bridge the gap between subjective literary criticism and natural language processing by creating a model that emulates the results of a survey into literary tastes. A panel of human experts qualified segments of literary text according to how aesthetically pleasing they found them. These segments were then rated for literariness in an open survey using a Likert scale. Each segment was subjected to a parts-of-speech tagger using NLTK and the results compared with those of the survey. Using a Grounded Theory approach, experiments using various combinations of parts-of-speech were carried out in order to build a model that could replicate the results shown in the open survey. The success of this approach confirms the feasibility of using this method to create a more accurate and analytical model of literary criticism involving deeper stylistic markers.