Enhancing Sentiment Analysis on Amazon Reviews Using Ensemble Learning Models

  • Unique Paper ID: 172458
  • Volume: 11
  • Issue: 8
  • PageNo: 3349-3354
  • Abstract:
  • With growing e-commerce, online review provides business with a vital source of insight into how to improve product quality and customer satisfaction. This paper presents a sentiment analysis of Amazon product reviews through advanced ensemble learning algorithms, which utilize Random Forest, Gradient Boosting, AdaBoost, and XGBoost classifiers of customer sentiments: positive, neutral, and negative. In contrast to other models like logistic regression and naive Bayes, ensemble methods yield better results due to the boosted decision trees. RandomizedSearchCV is employed as a technique to optimize hyperparameters that would produce better performance, and GPU acceleration ensures its computational efficiency. Models are evaluated based on accuracy, precision, recall, and confusion matrices, and results show that XGBoost outperforms the rest of the classifiers in the prediction of sentiment. This research emphasizes the fast-growing importance of machine learning in deriving insights out of large-scale textual data, hence helping out e-commerce platforms and businesses understand consumer preferences, improve recommendations, and adapt marketing strategies, demonstrating the magnitude of sentiment analysis in the enhancement of customer experience.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 11
  • Issue: 8
  • PageNo: 3349-3354

Enhancing Sentiment Analysis on Amazon Reviews Using Ensemble Learning Models

Related Articles