This project performs sentiment analysis on a dataset of Amazon reviews for wireless speakers from multiple brands (JBL, Bose, Sony, Ultimate Ears, and Beats). The goal is to analyze customer sentiments, identify key features driving positive and negative feedback, and visualize sentiment distributions to derive actionable insights. The analysis uses TextBlob for sentiment analysis, Pandas and NumPy for data processing, and Matplotlib, Seaborn, and WordCloud for visualizations.
The dataset includes reviews for products like JBL Flip 6, Ultimate Ears EVERBOOM, JBL Charge 5, Bose SoundLink Flex, Sony ULT Field 3, and Beats Studio Buds. The project follows a step-by-step approach to load the dataset, compute sentiment scores, categorize reviews, visualize results, and save the processed data.
- review_id: Unique identifier for each review
- product_id: Identifier for the product (e.g., B09GJVTRNZ for JBL Flip 6)
- title: Review title
- author: Reviewer name
- rating: Star rating (1–5)
- content: Review text (used for sentiment analysis)
- timestamp: Review date
- product_attributes: Product details (e.g., color, style)
- timestamp_cleaned: Cleaned review date in YYYY-MM-DD format
- Sentiment Distribution: Most reviews are positive (e.g., 80–90%), reflecting high star ratings (4–5 stars). Few reviews are neutral or negative.
- Polarity Histogram: Right-skewed, indicating predominantly positive sentiments due to high customer satisfaction.
- Polarity vs. Subjectivity Scatter Plot: Most reviews cluster in the positive polarity (>0) and high subjectivity (>0.5) quadrant, showing strong, opinionated praise.
- Word Cloud: Common terms in positive reviews include "sound," "bass," "quality," "battery," "portable," and "durability," highlighting valued features like sound quality and portability.
- Product-Specific Sentiment: Most products (e.g., JBL Flip 6, Bose SoundLink Flex) have overwhelmingly positive reviews. The Sony ULT Field 3 has some negative feedback (e.g., due to insufficient sound range).
