johnny12150/niser

Using SR-GNN to implement "NISER: Normalized Item and Session Representations to Handle Popularity Bias"

niser's Issues

Consider masking out the padding embeddiing at the tail of each session and adding positional embedding.

In the original paper the normalization is made with items in one session. While when implementing the algorithm we need to pad some positions to make the length of sessions the same in one batch. So when calculating the l2 norm the items at irrelative positions should be ignored. The author also found adding positional embedding is a little helpful. By the way, I'm looking forward for your number on Star GNN. Now I have reproduced the number for yoochoose1/64 while I can't get the same number on diginetica. Thank you a lot!

About paper

May I ask whether this paper has been published?

What specific methods are used in target-aware part of the code and is there any reference to this method?

I didn't find the description of the target-aware part of the code in the original article, but this part can really improve the effect of the model. Could you please tell me what specific methods are used in this part and is there any reference to this method? If so, I want to read it and further study it. Thank you very much.