Natural Language Processing-Based Literary Analysis Framework for Linguistic Nuance Detection in English Literature

The present paper introduces a Natural
Language Processing (NLP)-Based Literary Analysis
Framework that works to identify the linguistic elements
in English literature. The framework leverages state-of-the-art Contextual Word Embeddings in BERT
(Bidirectional Encoder Representations of Transformers),
a state-of-the-art NLP model to learn the semantic
nuances in literary works. Using BERT that allows
identifying contextual meanings of individual words
relying on the text around it, the framework can extract
such complex linguistic features as irony, metaphors, and
emotions in narratives. This method enables us to identify
sophistication in language used important phenomena that
cannot be identified in the conventional literary analysis,
thus a more insightful interpretation of the language and
style of an author. Implementation of BERT means that
we can identify literary devices, emotional changes and
thematic developments within a text which are incredibly
important in actually interpreting the text beyond the
generally accepted interpretations. Through the marriage
of computers and linguistics this framework finds a piece
of the puzzle as far as literary analysis. This framework
presents a new way of viewing English literature through
modern NLP techniques.