Search

Browse

• Communities
• Authors
• Year
• Titles
• Subjects

My Space

• Login

Intelligent sentiment analysis system using multicriteria and qualitative comparativeanalysis

Kyriakidis Anastasios

Full record

---

Summary

In today's dynamic and competitive market environment, understanding and enhancing customer satisfaction are essential for business success. Traditional methods of analysing customer feedback, often reliant on structured survey data, fall short in capturing the complexity and diversity of modern customer experiences reflected in heterogeneous data, such as those found in online customer reviews. The present dissertation proposes a novel five-step customer feedback analysis system designed to address these challenges by integrating advanced analytical techniques, including aspect-based sentiment analysis (ABSA), multicriteria decision analysis (MCDA), opinion analysis (OA), and qualitative comparative analysis (QCA).The system commences with the preprocessing of customer reviews to prepare data for subsequent analysis, followed by the application of an aspect-based sentiment analysis technique that leverages Large Language Models (LLMs) to extract and evaluate sentiment associated with specific aspects or features. The third step introduces the MUSAsent method, an extension of the Multicriteria Satisfaction Analysis (MUSA) framework, by incorporating sentiment data with quantitative ratings, providing a multi-dimensional analysis of customer satisfaction. The fourth step involves opinion analysis, which refines the satisfaction analysis through sentiment-based metrics, and the final step employs fuzzy-set Qualitative Comparative Analysis (fsQCA) to extract actionable insights and guidelines for strategic decision-making.The effectiveness of the system is demonstrated through experiments and case studies in the airlines and hospitality industries, showing its capability to translate complex customer feedback into practical, evidence-based strategies. By combining both quantitative and qualitative data, this dissertation contributes to the field of customer satisfaction analysis, by providing a robust framework for customer feedback analysis, offering businesses a powerful tool to enhance decision-making and improve customer satisfaction across diverse sectors.

Available Files

Services