Data Mining Group Project: Recommender System

Links

• Report
• Video Report
• Presentation Slides (though I'm absolutely sure you won't need it)
• Tableau

Developers info

Data collection and preparation

MovieLens 1M

python src/datasets/download_movie_lens.py - downloading MovieLens 1M dataset
python src/datasets/merge_movie_lens.py - preprocessing and merging MovieLens dataset

IMDB

python src/datasets/download_imdb.py - downloading IMDB dataset

Combination

To match titles in the movielens dataset and ids in imdb dataset, you should run

python src/datasets/extract_movielens_to_imdb_mapping.py

This creates data/movielens_to_imdb_mapping.csv file. For now, the script gives ~3000/3700 matches. The script may take ~10 minutes, so I suggest to use the existing matching file in from the repo.
After it, to extract IMDB features, run

python src/datasets/augment_movielens.py

The current version extracts average rating of the film and number of votes.