Skip to content

Latest commit

 

History

History
49 lines (39 loc) · 2.68 KB

README.md

File metadata and controls

49 lines (39 loc) · 2.68 KB

Movie Night Group Recommendation

Website

Visit: https://movie-night-recommendation.herokuapp.com
1. Note that it might take some time for the Heroku application to load, please be patient.
2. Note that it takes approximately 50 seconds for the recommendations to be returned.

Project setup

  1. git clone https://github.com/GroupMovieRec/group-movie-recommendation.git
  2. cd group-movie-recommendation
  3. python app.py
  4. Open browser at the respective address. (Chrome or Firefox recommended)
    For example, that address might look something like: http://0.0.0.0:5000/.

Testing setup

In the testing/ folder you'll be able to find all necessary files in order to run the tests for comparing our method's performance.

  • testing_group_differences.py and testing_method_comparison.py import nmf.py, rdfnmf.py, and rdfnmf_updated.py.
  • nmf.py is by created by Hung-Hsuan Chen [email protected].
  • rdfnmf.py and rdfnmf_updated.py are heavily modified by us, but are based on Hung-Hsuan Chen's work.
  • In the testing files you can change the 50 iterations to something less. When you create the models by initialising and declaring a class for the different methods you can also reset the n_epochs to something less than 50.

Acknowledgements

  • We would like to thank the authors of Movinder for publishing the code for their application, as a part of that has been the baseline for our user-interface. If you’re curious about the Movinder project, make sure to check them out their GitHub repository.
  • We would also like the thank Chen & Chen for publishing their code for the RDFNMF, as this was the baseline for our weighted matrix factorisation implementation. If you are interest in there work, check out their GitHub repository for their paper.

Project Report

The report for this project can be found here.

Information

Group project created in the context of TU Delft's CS4065 Multimedia Search and Recommendation.

Team 8:

  • Shreyan Biswas
  • Caroline Freyer
  • Francesca Drummer
  • Stefan Petrescu

Video presentation can be watched here.