Share CustomerInsight
🌟 Empowering businesses with AI-driven customer insights. Built for the next generation of data-driven decision making.
Tip
Experience the power of AI-driven customer analytics through our intuitive interface.
Main Dashboard - Comprehensive Analytics Overview
AI-Powered Analysis - Sentiment Detection and Keyword Extraction
📱 More Analytics Views
Topic Modeling and Clustering Analysis
Advanced Insights and Anomaly Detection
Tech Stack Badges:
Important
This project demonstrates modern NLP and machine learning practices with transformers, jieba, and streamlit. It combines AI-powered text analysis with interactive visualization to provide comprehensive customer insights. Features include sentiment analysis, keyword extraction, topic modeling, and anomaly detection.
📑 Table of Contents
- 📸 Project Screenshots
- 🌟 Introduction
- ✨ Key Features
- 🛠️ Tech Stack
- 🏗️ Architecture
- ⚡️ Performance
- 🚀 Getting Started
- 📖 Usage Guide
- 🔌 Data Format
- ⌨️ Development
- 🤝 Contributing
- 📄 License
- 👥 Team
- 🙋♀️ Author
We are passionate about transforming customer feedback into actionable business insights. By leveraging cutting-edge Natural Language Processing and Machine Learning technologies, CustomerInsight provides businesses with powerful, scalable, and user-friendly analytics tools.
Whether you're a business analyst, product manager, or data scientist, CustomerInsight will be your customer intelligence playground. Our platform specializes in multilingual text analysis with optimized support for Chinese and English languages.
Note
- Python 3.7+ required
- Streamlit for interactive web interface
- Pre-trained transformer models for sentiment analysis
- Jieba for Chinese text segmentation
- Plotly for interactive visualizations
 |
No installation required! Experience our platform firsthand. |
|---|
Tip
⭐ Star us to receive all release notifications and stay updated with the latest features!
Experience next-generation customer analytics with our comprehensive NLP suite. Our innovative approach provides unprecedented insights through advanced machine learning algorithms and transformer models.
Core Capabilities:
- 🎯 Sentiment Analysis: Advanced emotion detection with confidence scoring using BERT-based models
- 🔍 Keyword Extraction: TF-IDF and Jieba-powered keyword identification with trend analysis
- 🧠 Topic Modeling: LDA and K-means clustering for content categorization
- 🔬 Insight Analysis: Anomaly detection and correlation analysis for deep business insights
Supported Models:
- Chinese:
uer/roberta-base-finetuned-jd-binary-chinese - English:
nlptown/bert-base-multilingual-uncased-sentiment
Revolutionary data visualization that transforms how users interact with customer feedback. With our advanced charting capabilities and intuitive design, users can explore insights while maintaining clarity and actionability.
Visualization Types:
- 📊 Sentiment Trends: Time-series analysis with interactive filtering
- ☁️ Word Clouds: Dynamic keyword visualization with custom styling
- 🌐 Topic Networks: Network graphs showing content relationships
- 📈 Correlation Heatmaps: Statistical relationship visualization
Beyond the core analysis, CustomerInsight includes:
- 📁 Flexible Data Import: Support for CSV and Excel files with intelligent column mapping
- 🌐 Multi-language Support: Optimized for Chinese and English text processing
- 🔄 Real-time Processing: Batch analysis with progress tracking
- 📊 Interactive Dashboards: Streamlit-powered responsive interface
- 🎨 Customizable Visualizations: Plotly-based charts with export capabilities
- 💾 Data Export: Results export in multiple formats
- ⚙️ Configurable Parameters: Adjustable analysis settings for different use cases
- 🔍 Smart Filtering: Date range, rating, and category-based filtering
✨ More features are continuously being added as the project evolves.
Streamlit 1.2+ |
Python 3.7+ |
Pandas |
Plotly |
PyTorch |
Scikit-learn |
Core Framework:
- Frontend: Streamlit for interactive web applications
- Language: Python 3.7+ with type hints
- Data Processing: Pandas + NumPy for efficient data manipulation
- Visualization: Plotly for interactive charts and graphs
AI & NLP Stack:
- Deep Learning: PyTorch + Transformers for sentiment analysis
- Chinese NLP: Jieba for text segmentation and keyword extraction
- Machine Learning: Scikit-learn for clustering and statistical analysis
- Text Processing: NLTK for English language processing
Data & Visualization:
- Charts: Plotly Express for interactive visualizations
- Word Clouds: WordCloud with matplotlib integration
- Network Analysis: NetworkX for topic relationship graphs
- Data Export: Multiple format support (CSV, JSON, Images)
graph TB
subgraph "Data Input Layer"
A[CSV/Excel Files] --> B[Data Processor]
B --> C[Data Validation]
C --> D[Text Preprocessing]
end
subgraph "Analysis Engine"
D --> E[Sentiment Analyzer]
D --> F[Keyword Analyzer]
D --> G[Topic Analyzer]
D --> H[Insight Analyzer]
end
subgraph "AI Models"
I[BERT Models]
J[Jieba Segmentation]
K[TF-IDF Vectorizer]
L[LDA/KMeans]
end
subgraph "Visualization Layer"
M[Sentiment Visualizer]
N[Keyword Visualizer]
O[Topic Visualizer]
P[Insight Visualizer]
end
subgraph "Web Interface"
Q[Streamlit Dashboard]
R[Interactive Controls]
S[Real-time Updates]
end
E --> I
F --> J
F --> K
G --> L
E --> M
F --> N
G --> O
H --> P
M --> Q
N --> Q
O --> Q
P --> Q
customer-insight/
├── app.py # Main Streamlit application
├── src/ # Core analysis modules
│ ├── data_processor.py # Data loading and preprocessing
│ ├── text_analyzer.py # NLP analysis engines
│ └── visualizer.py # Visualization components
├── utils/ # Utility functions
│ ├── jieba_config.py # Chinese text processing
│ ├── text_cleaning.py # Text preprocessing
│ └── chinese_stopwords.txt # Chinese stopwords
├── public/ # Static assets
├── requirements.txt # Dependencies
├── setup.py # Installation script
└── example_dataset.csv # Sample data
Key Metrics:
- 🚀 Processing Speed: Handles 10,000+ reviews in under 2 minutes
- 💾 Memory Efficient: Optimized batch processing for large datasets
- 🎯 Accuracy: 90%+ sentiment classification accuracy
- 📊 Real-time Updates: Interactive analysis with progress tracking
Performance Optimizations:
- 🎯 Smart Caching: Streamlit caching for model loading and results
- 📦 Batch Processing: Efficient handling of large text datasets
- 🔄 Model Optimization: Pre-trained transformer models with GPU support
- 💨 Lazy Loading: On-demand analysis module initialization
Important
Ensure you have the following installed:
1. Clone Repository
git clone https://github.com/ChanMeng666/customer-insight.git
cd customer-insight2. Install Dependencies
# Install all required packages
pip install -r requirements.txt3. Setup Environment
# Run setup script to initialize environment
python setup.py install4. Launch Application
# Start the Streamlit application
streamlit run app.py🎉 Success! Open http://localhost:8501 to access CustomerInsight.
The setup script automatically:
- Creates necessary directories
- Downloads required models
- Configures Chinese stopwords
- Initializes Jieba dictionary
Manual Configuration (Optional):
# Custom jieba dictionary setup
import jieba
jieba.add_word('custom_term')
jieba.analyse.set_stop_words('path/to/stopwords.txt')Getting Started:
- Launch Application using
streamlit run app.py - Upload Data by dragging CSV/Excel files or using file browser
- Configure Analysis by selecting parameters and time ranges
- View Results through interactive dashboards and visualizations
Quick Analysis Workflow:
# Example data format
data = {
'content': ['Great product!', 'Poor quality', 'Excellent service'],
'rating': [5, 2, 5],
'timestamp': ['2024-01-01', '2024-01-02', '2024-01-03']
}Sentiment Analysis:
- Supports batch processing of thousands of reviews
- Confidence scoring for each prediction
- Historical trend analysis with time-series visualization
Keyword Extraction:
- TF-IDF based keyword identification
- Custom stopword filtering
- Trend analysis across time periods
Topic Modeling:
- LDA (Latent Dirichlet Allocation) for content categorization
- K-means clustering for similarity grouping
- Interactive topic exploration with examples
Insight Analysis:
- Anomaly detection for unusual patterns
- Correlation analysis between ratings and sentiment
- Statistical significance testing
Required Columns:
content: Review text contentrating: Numerical rating (1-5)timestamp: Date/time information
Optional Columns:
user_id: User identifiercategory: Product/service category
Supported Formats:
- CSV files with UTF-8 encoding
- Excel files (.xlsx, .xls)
- Headers in first row
Example Data Structure:
| timestamp | content | rating | user_id | category |
|---|---|---|---|---|
| 2024-01-01 | Excellent product quality! | 5 | user_001 | electronics |
| 2024-01-02 | Poor customer service | 2 | user_002 | support |
Setup Development Environment:
# Clone and setup
git clone https://github.com/ChanMeng666/customer-insight.git
cd customer-insight
# Install dependencies
pip install -r requirements.txt
# Start development server
streamlit run app.py --server.runOnSave trueDevelopment Scripts:
# Run tests
python -m pytest tests/
# Code formatting
black src/ app.py
# Type checking
mypy src/
# Lint code
flake8 src/ app.pyapp.py- Main Streamlit application entry pointsrc/data_processor.py- Data loading, validation, and preprocessingsrc/text_analyzer.py- Core NLP analysis classessrc/visualizer.py- Visualization componentsutils/- Helper functions and configurations
We welcome contributions! Here's how you can help improve CustomerInsight:
Development Process:
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Make your changes with proper tests
- Commit your changes (
git commit -m 'Add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
Contribution Guidelines:
- Follow Python PEP 8 style guidelines
- Add tests for new features
- Update documentation as needed
- Ensure all tests pass
Areas for Contribution:
- 🐛 Bug fixes and improvements
- 🌟 New analysis features
- 🎨 UI/UX enhancements
- 📚 Documentation improvements
- 🌐 Additional language support
This project is licensed under the MIT License - see the LICENSE file for details.
Open Source Benefits:
- ✅ Commercial use allowed
- ✅ Modification allowed
- ✅ Distribution allowed
- ✅ Private use allowed
Chan Meng Creator & Lead Developer |
Chan Meng
LinkedIn: chanmeng666
GitHub: ChanMeng666
Email: [email protected]
Website: chanmeng.live
Empowering data-driven decisions through AI-powered analytics
⭐ Star us on GitHub • 📖 Read the Documentation • 🐛 Report Issues • 💡 Request Features • 🤝 Contribute
Made with ❤️ by the CustomerInsight team
