COMPREHENSIVE PAPER ON EMOTIONAL STATE CLASSIFICATION IN POETRY

Emotional state classification in poetry is an emerging domain within artificial intelligence (AI), intersecting fields like natural language processing (NLP), affective computing, and literary analysis. Traditional sentiment analysis methods often struggle with poetry due to its layered emotions and figurative language. This survey examines the development of an attention-based Convolutional Bidirectional Long Short-Term Memory (C-BiLSTM) model specifically designed for emotion classification in poetry. Integrating CNN for local feature extraction, BiLSTM for sequential context, and attention mechanisms to highlight key emotional cues, the model captures nuanced emotional states such as love, joy, sadness, and anger. The paper discusses model architecture, training techniques, and evaluation methods, comparing C-BiLSTM’s performance with other machine learning approaches.