A system to scrape product’s feedback from website and classify them (bad or good) to suport customers in making choice ( buy or not)

Data was crawled from https://www.thegioididong.com/ Using 3 different models : RNN, SVM, NaiveBayes to classify and compare performance

For more details about training model, look at the notebook in 'Training' directory

Front-end

1. Install package:

Run command at ./phone-recommendation-frontend: yarn

2. Disable CORS in Chrome:

Run command: "C:\Program Files (x86)\Google\Chrome\Application\chrome.exe" --disable-web-security --disable-gpu --user-data-dir=~/chromeTemp

3. Run project:

Run command: yarn dev

Back-end

1. Install python:

https://www.python.org/downloads/

2. Install lib:

pip install Flask

pip install Flask-WTF

pip install underthesea

pip install scikit-learn==0.23.1

pip install keras==2.2.5

pip install TensorFlow==1.14

3. Run code:

py main.py