SentiMobile AI is an intelligent tool designed to decode customer sentiment on Amazon mobile phone reviews. Leveraging Natural Language Processing (NLP), this project goes beyond simple star ratings to understand the emotions behind the text—classifying feedback as Positive, Neutral, or Negative.
Developed by KhalidExe, this tool empowers users to visualize market trends and product reception through interactive data analytics.
- Smart Sentiment Detection: Uses
TextBlobto assign polarity scores to thousands of reviews instantly. - Interactive Dashboard: Run the notebook and select any of the top 20 smartphones to analyze.
- Visual Insights: Generates dynamic Pie Charts and Bar Graphs to showcase user satisfaction.
- Automated Reporting: Exports the analyzed results automatically to a CSV file for further use.
- Core: Python
- NLP Engine: TextBlob
- Data Science: Pandas, NumPy
- Visualization: Matplotlib
- Interface: Colorama (for styled terminal outputs)
This project uses the Amazon Unlocked Mobile Phones dataset from Kaggle.
- Download the dataset: Click here to download from Kaggle.
- Setup: Unzip the downloaded file and place
Amazon_Unlocked_Mobile.csvin the main project folder. (Note: The dataset file is not included in this repo to keep it lightweight.)
- bash
git clone https://github.com/KhalidExe/SentiMobile-AI.gitcd SentiMobile-AI
- Bash
pip install -r requirements.txt
Ensure the CSV file is in the directory, then launch the notebook:
- Bash
jupyter notebook Sentiment_Analysis.ipynb
Input: The system reads raw Amazon review data.
Process: You select a product (e.g., iPhone 5s or Samsung Galaxy).
Analysis: The AI engine processes the text, removing noise and calculating sentiment scores.
Output: You get a detailed breakdown of what customers actually think (Satisfaction vs. Frustration).
Contributions are welcome! If you have ideas to improve the accuracy or add new features, feel free to fork the repo and submit a Pull Request.
KhalidExe
Field: Artificial Intelligence (AI)
GitHub: KhalidExe