<![CDATA[Sentiment analysis on marketplace customer reviews is important for understanding user perceptions and supporting Customer Relationship Management (CRM) strategies. This study proposes a multi-source sentiment analysis approach based on big data from Shopee and Tokopedia platforms using Hybrid Machine Learning. The research process includes data collection, preprocessing, TF-IDF feature extraction, and classification using Support Vector Machine (SVM) and Random Forest techniques. The preprocessing stage consists of case folding, tokenization, stopword removal, and stemming to improve the quality of textual data. The TF-IDF method is used to transform text data into numerical features before classification. The evaluation results show that the SVM model achieved an accuracy of 97.49%, while the Random Forest model achieved 97.47%. The sentiment distribution indicates a strong positive bias, reflecting high customer satisfaction with marketplace services. However, negative sentiment persisted, mainly due to delivery delays, application errors, and customer service issues. The proposed hybrid approach can provide data-driven insights to improve service quality and support decision-making in CRM strategies.]]>
Copyrights © 2026
