Using demographics and topic distributions of Facebook groups this projects attempts to see which parts can be classified using Random Forest models.
| File | Description |
|---|---|
| random_forest_all_predictors_v3.py | Script performing Random Forest model with varying dependent and independent predictors without grid search |
| random_forest_finalfeatures_v2.py | Models with most important features |
The file used as input can be found at grundtvig and is called . It contains 13672 rows with data from Facebook groups. The columns used in the models in these scripts are described below
| Column | Description |
|---|---|
| total_unique_p_c | amount of unique users that either comment or post in a facebook group |
| total_post | total amount of posts |
| total_comment | total amount of comments |
| dominance | gender dominance calculated with a 20/80 threshold (if males > 80% = 1, if females > 80 % = 2, else neutral = 3) |
| new_days | group age / longevity of group in days |
cd FB-gender-classification # Change directory to the repository folder
bash create_venv.sh
source FB_RANDOMFOREST/bin/activate
Run script by typing
python3 random_forest_all_predictors_v3.py # Run script
To deactivate and remove the environment, the following commands need to be executed:
deactivate
bash kill_vision_venv.sh