https://github.com/phenomsg/restaurant-review-analysis
This project utilizes the power of BERT (Bidirectional Encoder Representations from Transformers) for sentiment analysis
https://github.com/phenomsg/restaurant-review-analysis
bert-model data-science huggingface natural-language-processing python transformers
Last synced: 3 months ago
JSON representation
This project utilizes the power of BERT (Bidirectional Encoder Representations from Transformers) for sentiment analysis
- Host: GitHub
- URL: https://github.com/phenomsg/restaurant-review-analysis
- Owner: PhenomSG
- License: mit
- Created: 2024-01-17T11:15:02.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-07-18T20:28:33.000Z (11 months ago)
- Last Synced: 2024-08-06T08:11:03.505Z (10 months ago)
- Topics: bert-model, data-science, huggingface, natural-language-processing, python, transformers
- Language: Jupyter Notebook
- Homepage:
- Size: 53.8 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Restaurant Review Analysis System
## Project Overview
This project aims to build a comprehensive system for collecting and analyzing restaurant reviews. The system will:
1. Collect reviews from various platforms.
2. Analyze the sentiment of the reviews to generate a rating.
3. Provide explanations for the generated ratings.
4. Create a word cloud highlighting key themes from the reviews.
5. Analyze restaurant images to classify them based on suitability for different dining experiences (e.g., family, casual, date) and suggest improvements.
## Features
- **Review Collection**: Gather restaurant reviews from sources like Yelp and Google Reviews.
- **Sentiment Analysis**: Analyze the sentiment of the reviews to derive a rating.
- **Rating Explanation**: Provide detailed explanations for the generated ratings.
- **Word Cloud Generation**: Identify common themes in the reviews to create a word cloud.
- **Image Analysis**: Classify the restaurant type based on images and suggest improvements.
## Technology Stack
- **Python**: Main programming language.
- **Gemini Pro API**: Used for NLP and image analysis tasks.
- **NLP Libraries**: For additional natural language processing.
- **Image Processing Libraries**: For analyzing restaurant images.
- **Web Scraping Tools**: To collect reviews from various platforms.
## Installation
### Prerequisites
- Python 3.7 or higher
- API key for Gemini Pro
### Setup
1. Clone the repository:
```bash
git clone https://github.com/PhenomSG/Sentiment_Analysis_Using_NLP.git
cd restaurant-review-analysis
```
2. Create and activate a virtual environment:
```bash
python -m venv venv
source venv/bin/activate # On Windows use `venv\Scripts\activate`
```
3. Install the required packages:
```bash
pip install -r requirements.txt
```
4. Set up your Gemini Pro API key:
```bash
export GEMINI_API_KEY='your_gemini_api_key'
```
## Usage
### Collecting Reviews
To collect reviews from a source (e.g., Yelp), use the provided `collect_reviews_from_source` function:
```python
from review_collector import collect_reviews_from_source
reviews = collect_reviews_from_source('yelp', 'restaurant_id')
```
### Analyzing Reviews
To analyze sentiment and generate a rating:
```python
from review_analysis import sentiment_analysis, aggregate_ratings, gemini_pro_explain_rating
ratings = sentiment_analysis(reviews, model='gemini_pro')
overall_rating = aggregate_ratings(ratings)
explanation = gemini_pro_explain_rating(reviews, overall_rating)
```
### Generating a Word Cloud
To generate a word cloud from the reviews:
```python
from word_cloud_generator import extract_keywords, generate_word_cloud
word_cloud_data = extract_keywords(reviews)
generate_word_cloud(word_cloud_data)
```
### Analyzing Images
To classify restaurant type and suggest improvements based on images:
```python
from image_analysis import classify_restaurant_type, assess_image_quality
for image in restaurant_images:
category = classify_restaurant_type(image, model='gemini_pro')
improvements = assess_image_quality(image, model='gemini_pro')
print(f"Category: {category}, Improvements: {improvements}")
```
## Directory Structure
```
restaurant-review-analysis/
│
├── README.md # Project overview and instructions
├── requirements.txt # List of dependencies
├── collect_reviews.py # Functions for collecting reviews
├── review_analysis.py # Functions for analyzing reviews
├── word_cloud_generator.py # Functions for generating word clouds
├── image_analysis.py # Functions for analyzing images
└── main.py # Main script to run the analysis
```
## License
This project is licensed under the MIT License. See the [LICENSE](LICENSE) file for details.