University of Leuven – Machine Learning Project
Make sure you are running Python 3.
Run pip install -r requirements.txt
.
Add and enable the user script urlStreamHandler.user.js
in your browser. (Using Greasemonkey in Firefox, for example).
Run python urlStreamHandler.py
.
Now after a while, when you are on a page that you have already visited, the app will start suggesting pages you might want to go to.
If you want to pre-train the app with historical web usage data, call for example:
python urlStreamHandler.py --csv log1.csv log2.csv
.
The preprocessing and prediction code can be found in url_predictor.py
.
The code used to validate the model can be found in model_validator.py
.
The hours we worked on this project can be found here.