Unmasking Reviews: Analyzing Sentiments on Levoit’s Air Purifier Reviews on Amazon

        A data analytics capstone project that focuses on extracting and analyzing customer reviews of Levoit air purifiers from Amazon using web scraping techniques via DataMiner. The collected verified reviews underwent data preprocessing such as cleaning, tokenization, stop-word removal, and lemmatization to prepare the dataset for analysis. Sentiment classification was performed using VADER, a lexicon-based sentiment analysis tool, and enhanced through machine learning models including Logistic Regression, Naive Bayes, and Random Forest.

​

       The Logistic Regression model achieved the highest performance and was selected as part of a hybrid approach for more accurate sentiment prediction. The results were visualized through an interactive web-based dashboard featuring donut charts, word clouds, gauge charts, and review insights, providing a clear understanding of customer sentiment and product perception.