Computational Poetics is the scientific perspective that analyses literary texts quantitatively by help of state-of-the-art NLP and machine learning methods.
Its goal is to understand which text features drive the fluent, immersive or dysfluent, reflective reading of narratives and poetry (cf. Jacobs, 2015, 2018).
It focuses on affective-aesthetic features (e.g., Emotion, Immersion and Aesthetic Potential) in their dynamic interaction with cognitive aspects (e.g., processing fluency, syntactic complexity, readability; cf. Jacobs et al., 2017).
Key features are analysed at different text levels (e.g., phonological, semantic) and for different types (e.g., sublexical, interlexical features) based on the 4×4 matrix especially designed for ACTA shown in the left-hand figures (cf. Jacobs, 2015, 2018a).
The following pages present example applications of computational poetics studies from Emotion Potential to Sound Beauty.
