Leveraging Bernoulli Naïve Bayes (BNB) and Support Vector Machine (SVM) for Sentiment Analysis and Visualization of Sayurbox Application Reviews

User ratings on sites like the Google Play Store have grown to be a major factor in determining how successful an app is in the quickly changing digital world. These user experiences and opinions-rich reviews are a useful tool for sentiment analysis. We employ the Bernoulli Naive Bayes model as it is easy to apply, effective, and efficient in handling large, highly dimensional datasets-as text data like user reviews frequently has. Working on the assumption of conditional independence given the class label, the model often yields surprisingly good results even if it appears overly simplistic for natural language data. Our study attempts to display and extract sentiments from Google Play Store reviews, therefore offering insightful information about user experiences and impressions. The model successfully predicted the review test data as positive sentiment by 30.57% or 70 reviews and negative sentiment by 69.43% or 159 reviews, on the other hand. The Support Vector Machine model shows better accuracy of 74% when using Sentiment Analysis, compared to Naive Bayes with an accuracy of 73%. Our study will, we believe, significantly advance the academic discipline of sentiment analysis and have practical implications for marketers, software developers, and corporate decision-makers. Our work opens doors for more study in this fascinating area by demonstrating a promising next step toward using machine learning for sentiment analysis.