ABDUL RASHEED P , DR. SAYED FIROJ ALLI , DR. SOMA PARIJA , DR. JUNMONI BORGOHAIN , DR. S. MUNIRA BANU , SUSHIL DOHARE
DOI: https://doi.org/Emotional expression shapes narrative meaning and character psychology across literary fiction, yet traditional literary analysis relies heavily on human interpretation, limiting scalability, consistency, and empirical depth. Advances in computational psycholinguistics and natural language processing now offer powerful tools to analyze emotional cues, thematic patterns, and psychological states embedded within character discourse. This study develops an AI-driven framework for decoding literary emotions, leveraging sentiment analysis, cognitive-affective lexicons, semantic embeddings, and transformer-based contextual models. Using a curated corpus of English fiction spanning modernist, postmodern, and contemporary narrative traditions, the research investigates how linguistic markers signal emotional shifts, psychological complexity, and narrative tension. Results reveal strong correlations between psycholinguistic features and character roles, narrative arcs, and conflict structures. Emotional granularity such as guilt, longing, suspicion, or suppressed anger emerges through semantic drift, valence shifts, and contextual dependencies detectable through advanced NLP pipelines. While computational models enhance interpretive precision and uncover hidden emotional structures, limitations include figurative ambiguity, symbolic language, and culture-specific metaphors. The findings demonstrate that AI-enhanced literary interpretation offers a transformative pathway for narrative psychology, digital humanities, and computational literary studies.
