A Combined Sentiment and Statistical Analysis for Netflix User Reviews

This study aimed to analyze user perceptions of the Netflix mobile app by combining sentiment analysis and statistical modeling. A total of 13,500 reviews were scraped from the Google Play Store using Python, then processed and analyzed in R. Sentiment scores were computed using the Syuzhet package, and key patterns were visualized through word clouds and emotion plots. Spearman correlation, ordinal logistic regression, and multinomial logistic regression were used to examine relationships between sentiment, star ratings, and thumbs-up counts. Results showed that 61.89% of reviews were positive, with “good,” “love,” and “awesome” as dominant emotions. Sentiment scores were strong predictors of user ratings, while thumbs-up counts were slightly associated with critical reviews. In conclusion, combining sentiment and statistical methods provides deeper insight into user satisfaction, with emotionally detailed reviews—both positive and negative—drawing more engagement.