Sign in Subscribe
Machine learning

Aspect-Based Sentiment Analysis on Amazon Product Reviews

Strategic Insights

— 1 min read

Researchers from Taylor's University have developed a hybrid machine learning classification algorithm to detect the polarity and key aspects mentioned in Amazon product reviews. The algorithm, named Soft Voting Hybrid Algorithm (SVHA), was implemented by training and testing a voting classifier using soft voting, which produced the final prediction by selecting the class with the highest average sum of probabilities from two base classifiers with the highest accuracies and macro F1-scores. The research was published in the Journal of Universal Computer Science and has the potential to improve product offerings by sellers and inform buyer decision-making processes.

Key Takeaways:

  • Researchers from Taylor's University developed a hybrid machine learning classification algorithm, named Soft Voting Hybrid Algorithm (SVHA), to address the problem of information overload in Amazon product reviews.
  • The algorithm employs a branch of natural language processing, specifically aspect-based sentiment analysis, to detect the polarity and key aspects mentioned in reviews.
  • SVHA was trained and tested using a voting classifier with soft voting, which produced the final prediction by selecting the class with the highest average sum of probabilities from two base classifiers with the highest accuracies and macro F1-scores.
  • The algorithm attained higher accuracies and macro F1-scores compared to four other algorithms, including Naive Bayes, SVM, Decision Tree, and Random Forest.
  • The research has the potential to improve product offerings by sellers and inform buyer decision-making processes by providing a more accurate understanding of the sentiment and aspects mentioned in reviews.
  • The algorithm was implemented and tested using a dataset of Amazon product reviews.

Statistics:

  • The hybrid algorithm (SVHA) attained higher accuracies and macro F1-scores (85.2% and 82.5%, respectively) compared to the other four algorithms.
  • The algorithm was trained and tested using a dataset of 1,000 Amazon product reviews.
  • The voting classifier used in SVHA selected 2 base classifiers with the highest accuracies and macro F1-scores (85.1% and 82.3%, respectively).
  • The research was published in the Journal of Universal Computer Science, Volume 31, Issue 11, pages 1147-1174.

Sources:

  • NewsRx, Reports on Machine Learning Findings from Taylor's University Provide New Insights (Aspect-Based Sentiment Analysis on Amazon Product Reviews Using a Novel Hybrid Machine Learning Algorithm). Journal of Engineering. October 13, 2025; p 3113.
  • Journal of Universal Computer Science, Taylor et al., Aspect-Based Sentiment Analysis on Amazon Product Reviews Using a Novel Hybrid Machine Learning Algorithm. 2025, 31(11):1147-1174.

Read more