99731 / serec Goto Github PK

Hello. I am a master's student and studying recommendation systems. These days, I am studying your paper, SERec, and I have some questions. The questions are below.

1. On line 690 of socialexpo.cpp, I think this code asks for gradient b. Therefore, I think gamma_i doesn't have to be here. Is it right?

2. In paper, there is not comment about data preprocessing. Is it right that you used delicious, one of the datasets, as it was in the paper and no preprocessing?

3. When question 2 is right, I used delicious as a dataset and set hyperparameters as your paper. Then, I got a result with your code, but result is not good. Could you help me?

I would really appreciate it if you could answer my questions.
Thank you.