In this project, I developed a Naive Bayes classifier to predict the ratings of Amazon customer reviews based solely on their textual content. By analyzing patterns in the review text, the model estimates whether a review corresponds to a 1- to 5-star rating, providing insights into customer sentiment.
Data Preprocessing: Cleaned and prepared the dataset by handling missing values and converting review timestamps to a standardized format.
Feature Extraction: Utilized CountVectorizer to transform review texts into a bag-of-words representation, facilitating effective text analysis.
Model Training: Implemented a Multinomial Naive Bayes classifier to learn the relationship between review text and corresponding ratings.
Model Evaluation: Assessed the classifier's performance using metrics such as accuracy, precision, recall, and F1 score to ensure reliable predictions.
Visualization: Created visual representations, including histograms and pie charts, to illustrate the distribution of ratings and the proportion of verified reviews.