This application is a recommender system uses collaborative filtering
- Summary of set up
Download the file, extract it
Run
sudo python3 webserver.py
to start the server
Run
sudo python3 webclient.py
to run the client, enter CustomerID and it will return a list of ProductCode for recommendations, default maximum is 5
- Configuration
To configure the maximum item to introduce, one can either modify
recommender.Recommender.recommend(self,name,number=5)
change the number=5 to any integer u want
As this recommender use knn, the k can be configure by modifying 'recommender.Recommender.kneighbors(self,name,k=5)`
- Dependencies
This application depends on
- Numpy
- Scipy
- Pandas
- sklearn