Over the past decade, sentiment analysis has emerged as a pivotal tool for interpreting customer feedback and shaping service delivery strategies. With the explosion of user-generated content across digital platforms, organizations face both an opportunity and a challenge: extracting meaningful, emotion-rich insights from vast, unstructured data. This research investigates the evolving landscape of sentiment analysis in customer feedback systems, with a focused lens on emotion-centric service optimization. Through a comprehensive review of literature published from 2014 to 2024, we identify key advances in natural language processing (NLP), including machine learning and deep learning-based approaches that have enhanced the accuracy of sentiment detection. The study explores frameworks that move beyond basic polarity classification, aiming instead to map nuanced emotional states such as frustration, satisfaction, or anxiety. Furthermore, we analyze how sentiment-driven models have been integrated into real-time customer experience management systems to personalize interactions, reduce churn, and foster brand loyalty. Our findings reveal a trend toward hybrid models combining rule-based methods with contextual deep learning architectures, significantly improving interpretability and domain adaptability. We propose a sentiment-to-action framework that enables businesses to translate customer emotions into measurable service improvements. This research underscores the strategic value of emotion-aware sentiment analysis in delivering responsive, human-centered experiences and provides a roadmap for future developments in intelligent feedback systems.

Keywords: Sentiment Analysis, Customer Feedback, Emotion-Centric Optimization, Natural Language Processing (NLP), Customer Experience Management, Machine Learning