Model for one Recommender System
| Movielens 1M | RMSE | MAE |
|---|---|---|
| NormalPredictor | 1.5037 | 1.2051 |
| BaselineOnly | .9086 | .7194 |
| KNNBasic | .9207 | .7250 |
| KNNWithMeans | .9292 | .7386 |
| KNNBaseline | .8949 | .7063 |
| SVD | .8738 | .6858 |
| NMF | .9155 | .7232 |
| Slope One | .9065 | .7144 |
| Co clustering | .9155 | .7174 |
| Word2Vec | .8761 | .7168 |
*table is modified from origin
*Original: https://github.com/NicolasHug/Surprise
-
Get the data:
$ wget http://files.grouplens.org/datasets/movielens/ml-1m.zip && mkdir -p ~/data/movielens && unzip -d ~/data/movielens/ ml-1m.zip && rm ml-1m.zip -
Download and install Elasticsearch (tested on 5.2 and 6.3) https://www.elastic.co/guide/en/elasticsearch/reference/current/_installation.html
-
Install dependencies
$ pip3 install gensim sklearn tqdm jupyter pandas numpy
Vualaá! Now You are ready to run the code.