Fake Review Detection Project

Set up environment

1. Create virtual environment:

virtualenv -venv

2. Activate virtual environment:

./venv/Scripts/activate

3. While virtual environment is activated, install required modules

pip install -r requirements.txt

Load Dataset

1. Download YelpCSV folder, dataset_v1.csv, and review_feature_table.csv into DATA subdirectory. *Note: Program can be executed without downloading dataset_v1.csv and review_feature_table.csv beforehand. This will force program to re-compute all features and conduct undersampling, which will add to overall execution time.

.
├── DATA
     └── YelpCSV
            └── metadata.csv
            └── productIdMapping.csv
            └── reviewContent.csv
            └── reviewGraph.csv
            └── userIdMapping.csv
     └── dataset_v1.csv
     └── review_feature_table.csv

Execute program

1. Make sure your virtual environment is activated AND your current working directory is the general repo directory (ex. path/to/dir/fake_review_detection) and run command:

python ./CODE/main.py

After Executing

All necessary metrics and visualizations will populate in EVALUATIONS subdirectory.