Fake-Review-Detection

Detecting Fake Reviews using Semi-Supervised Learning from the Yelp Restaurant Reviews Dataset

Directory Structure

• Code

  • main.py --> Main Python File containing the code for the entire project
  • FakeReviewDetection.sh --> Script File to Run main.py. (This file installs all the libraries required for running the project using pip3 and runs the code using python3. If using a different version, please change the command from pip3 to pip/pip2 and python3 to python depending on your version.)

• Data

  • df.csv --> This is a cleaned, pre-processed and feature engineered version of the original dataset. (Provided for reference)

• Evals --> Contains Screenshots of various outputs and test runs of both the ML models.

How to Run

1. Clone the github repository onto your desktop.
2. Download the original dataset from the the following link (936MB). (Unable to upload this file as Github does not support uploading large files.)
3. Copy the dataset .db file in the following directory. Fake-Review-Detection/Data
4. Open terminal inside the Code directory.
5. Enter the following command to give permissions to FakeReviewDetection.sh file. chmod 777 FakeReviewDetection.sh
6. Run the script file using the following command. ./FakeReviewDetection.sh