TSKS11 course project

This project is published here only for the sake of sharing the knowledge. Unfortunately, the data that was used in the program could not be distributed beyond the participants of the TSKS11 course. To show a vague idea of how the test files looked like, please see below:

$ ls -hog data
total 4,7M
-rw-r--r-- 1  66K kwi 24 16:02 baseline.moviename
-rw-r--r-- 1 228K kwi 24 16:02 baseline.test
-rw-r--r-- 1 2,1M kwi 24 16:02 baseline.training
-rw-r--r-- 1  66K kwi 24 16:02 improved.moviename
-rw-r--r-- 1 236K kwi 24 16:02 improved.test
-rw-r--r-- 1 2,1M kwi 24 16:02 improved.training

All files originaly placed under data/ directory (this program also requires that).

• *.training and *.test - (user, movie, rating): three numbers per row, comma-separated integers, connecting the indices user-movie with the rating that the user gave for the movie. As an example, row 4,1,4 indicates that user 4 gave the movie 1 a rating of 4.
• *.moviename - (index; title; additional info): for example: 1; Toy Story (1995); Animation|Children's|Comedy

pipenv install

pipenv run python recommend.py {baseline|improved}