This is the implementation for our paper "Learning and Fusing Multiple User Interest Representations for Micro-Video and Movie Recommendations."
Python 3.6, Tensorflow 1.12
Since the limited space in GitHub, we upload the Movie data(password: ah6i) and Micro-video data(password: jp82) on BaiduYun for running our code. You could run mkdir Files, download the dataset, and then put them into Files/.
- For Micro-video Recommendations, run
cd Video/LIN/LIN_Video;python launcher.py;
cd Video/IIN/IIN_Video;python launcher.py;
cd Video/CIN/CIN_Video;python launcher.py;
cd Video/NIN/NIN_Video;python launcher.py;
Finally, you can perform late fusion to get final prediction of MUIR for micro-video recommendations.
- For Movie Recommedation, run
cd Movie/LIN/LIN_Movie;python launcher.py;
cd Movie/IIN/IIN_Movie;python launcher.py;
cd Movie/CIN/CIN_Movie;python launcher.py;
cd Movie/NIN/NIN_Movie;python launcher.py;
Finally, you can perform late fusion to get final prediction of MUIR for movie recommendations.