Sentiment Analysis of Customer Feedback and Reviews in E-Commerce Systems

Natural Language Processing (NLP) is an effective tool for analyzing consumer satisfaction with products and services typically sold by e-businesses. This study explores the use of sentiment analysis to understand customer feedback and ratings using a dataset acquired by Amazon. That dataset covers online purchases over a period of eight years. The study aims to examine the factors that impact sentiment polarity throughout different periods and understand changes in consumer feelings and reviews towards online products and services. The applied method comprehensively studies sentiment, extracting features using the Term Frequency-Inverse Document Frequency (TF-IDF) and other text analysis and data processing procedures. Visualization is also applied to illustrate the trends, patterns, and changes in customer satisfaction over time and per product and line of products. The findings underscore the importance of sentimental research in discerning customer sentiments on online e-commerce platforms and offer practical recommendations for enhancing online and improving customer satisfaction.