Sentiment Analysis on Ibsen’s “A Doll’s House”

Abstract

This study applied sentiment analysis to Ibsen’s “A Doll’s House” to investigate the potential of deep learning-based sentiment analysis in examining the underlying structure of modern drama and to explore optimal strategies for its practical application. Our exploration results underscore the potential of sentiment analysis as a methodology for analysis in literary studies. We utilized three distinct measures to process sentiment scores: mean sentiment scores, moving average sentiment curves, and cumulative sentiment curves. Each of these measures consistently resonated with the play’s themes and content, thereby underscoring their relevance in literary studies. Specifically, mean sentiment scores proved beneficial in encapsulating the overall sentiment profiles of the characters. Moving average sentiment curves excelled in tracing the dynamic fluctuations of sentiment throughout the narrative. Lastly, cumulative sentiment curves offered a comprehensive perspective of sentiment trends across the play. Despite these encouraging findings, the study also highlights the necessity for more refined and context-specific models and techniques for a more accurate and detailed sentiment analysis in literature.