mkmksto / ln-recommender Goto Github PK

Open in Colab and follow the instructions.

• Create a directory and unpack all the contents of the contents of the zip, ex. ln-recommender
• Create python venv

cd ln-recommender
python -m venv venv

• Activate your venv depending on your OS

venv\Scripts\activate{.bat|ps1}

• Install package and its dependencies

pip install .

• Run the tool

ln-recommender -h

• Train model and output model file

ln-recommender train -td data.csv -ms model.cbm

• Train model with evaluation dataset and output model file, can yield better results if you have lots of data

ln-recommender train -td data.csv -ms model.cbm -ev

• Evaluate single text file

ln-recommender eval -t text.txt

• Evaluate directory with a saved model

ln-recommender eval -d directory -ml model.cbm

• Evaluate directory with a saved model and save output to csv file

ln-recommender eval -d directory -ml model.cbm -o out.csv

• Run clustering with minimum group size 3 and core samples 2

ln-recommender cluster -td data.csv -mcs 3 -ms 2