Skip to content

Latest commit

 

History

History
36 lines (28 loc) · 1.76 KB

README.md

File metadata and controls

36 lines (28 loc) · 1.76 KB
Raw

#Movie Recommender (Flask) Build Status

##Description: This is a very simple web app built in Flask, which uses the MoveLens dataset of 100k user ratings of popular movies in order to make recommendations on what movies a given user may want to watch. It was a quick(ish) weekend experiment to make a web-driven interface to explore pure Python objects, with no database on the backend. As a result, it will take a couple minutes to load the first time, in order to build the network of recommendations.

##Requirements:

  • Python 3 (3.4 recommended)
  • MovieLens 100k dataset (ml-100k.zip), unzipped into ./datasets/ml-100k
  • Flask
    • Install via pip3 install -r requirements.txt

##Instructions:

  1. Download/clone all files from this repo to your machine
  2. Download and unzip the required dataset, described above
  3. Run on the command line: "python3 ui.py"*
  4. Open http://127.0.0.1:5000/ in a web browser window
  5. Explore the site to find ratings for movies, movie rankings, and recommendations for each user based on their previous ratings

######* Note: This application is under development, and uses debug mode and as such, should not be used in a production environment!

####Acknowledgements: The MovieLens data sets used for this project were collected by the [GroupLens Research Project] (http://grouplens.org) at the University of Minnesota.

This work is neither endorsed nor sponsored by the University of Minnesota nor the GroupLens Research Group.

This project and the provided data are for non-commercial research purposes.