The purpose of this project is to implement the methods proposed by Yu Rong, Xiao Wen, and Hong Cheng at WWW '14, and if possible, extend it. We will be using the Movie Lens 1M Dataset to test the results of the method, as per the paper.
This project uses Python 3, and uses NumPy as a dependency.
Majority of the implementation is done on the Jupyter Notebook, after which it was ported to a Python document to form a coherent class structure. You can use our improved algorithm by passing alternative = True to the Predicter class initializer in paper.py.