tafoca/product-rating

Product Rating system using sentiment analysis of text reviews.

Product Rating System

Using Deep Learning

---

About

This Project generates rating of a product in an e-commerce website by analyzing the text based review provided by the users using deep learning.

The system uses sequence learning methodology in order to achieve desired functionality. The reviews provided by the users about a product on an e-commerce website is analyzed and rating of that product is generated based on the review. This system leverages the power of machine learning to completely eradicates the trouble of giving rating as well as writing review and helps to predict accurate rating based on user reviews.

Learn More

Installation

The application runs on Python's Flask micro web framework. The following packages are required to be installed in order to run this application.

Packages for Sentiment Analysis

• numpy
• pandas
• tqdm
• tensorflow
• sklearn
• matplotlib
• keras
• nltk

Packages for Web Application

• flask
• SQLAlchemy
• Flask-MySQL
• mysqlclient
• mysql-connector-python

Install these packages using pip (i.e. pip install numpy)

If problems arises during installing mysqlclient (eg: Microsoft Visual C++ 14.0 is required), visit this link and download .whl file for your system and python version and run
pip install c:\location_to_whl_file\mysqlclient‑1.3.13‑cp36‑cp36m‑win_amd64.whl

Download LSTM Model and Text Data frome here and place them inside the folder containing above files.

The folder should look as follows:

Database

• Create a database named rating in your SQL Server (database and table names can be changed inside rating.sql file).
• Import the rating.sql into your SQL Database.
• Use Xampp DB manager

Usage

Open Command Prompt (CMD) and navigate to the folder containing the files. Then run flask run OR python app.py

Visit http://127.0.0.1:5000 in your browser to use the application.

Contributors

• Pukar [Website]
• Rajesh Pandey
• Dependra Ramtel
• Laurent Tabueu Fotso