A Novel Narrative Analysis Approach for Digital Content Using Natural Language Processing

In today’s digital world, the unprecedented volume of online content is shaping global narratives, influencing opinions, policies, and cultures in real-time. Many researchers have proposed narrative analysis algorithms, but they cannot automate the content selection and require manual interventions. Another key issue is the limited applicability across diverse domains, leading to inefficiencies and biases in handling large, unstructured datasets. There is a significant need to develop novel approaches to enable narrative analysis from vast amounts of available digital content for better decision-making, analysing social trends, and addressing misinformation and bias. This paper uses the abstractive summarisation approach to improve narratives’ identification, selection, analysis, and presentation. The research aims to develop a narrative analysis algorithm using natural language processing techniques to automate the selection of digital media content, improve accuracy, and reduce bias. The proposed approach shows an $87 \%$ accuracy in story arc identification with discourse enhancement, while significantly reducing the time required for content selection and analysis.