##"Using Machine Learning for Creating a Movie Recommendation System"
My goal in this project is to use machine learning to be able to build a movie recommendation system. I was able to create four models; a decision tree classifier, a K nearest neighbor classifier, a random forest classifier and a LightGBM classifier.
The performance of the models is compared to a benchmark model which is the user based collaborative filtering. The K nearest neighbor classifier was able to produce a higher F1 score higher than the benchmark model.
This project requires Python 2.7(if you complete this project in Python 3.x, you will have to update the code in various places including all relevant print statements) and the following Python libraries installed:
For the LightGBM Model you will need to:
-
Install LightGBM using: conda install -c conda-forge lightgbm
-
import lightgbm as lgb in your code.
You will also need to have software installed to run and execute a Jupyter Notebook
If you do not have Python installed yet, it is highly recommended that you install the Anaconda distribution of Python, which already has the above packages and more included. Make sure that you select the Python 2.7 installer and not the Python 3.x installer.
Template code is provided in the movierecommender.ipynb notebook file.
The MovieLens dataset (ml-latest-small) [5] to be used in this capstone project will be taken from ”https://grouplens.org/datasets/movielens/“ the dataset describes a 5-star rating and a free-text tagging activity from a movie recommendation service called “MovieLens”.
Features
userId:userId represents a unique key for each user.movieId: movieId represents a unique key for each user.genres:list of the different movie genres separated by '|'.title: contains the movie name in addition to the production year.
Target Variable
4. rating: Ratings of movies out of 5, with half-star increments.