search

HKUDS / AdaGCL

[KDD'2023] "AdaGCL: Adaptive Graph Contrastive Learning for Recommendation"
https://arxiv.org/abs/2305.10837
54 stars 2 forks source link

Question in the dataloader #6

Closed fanglin1 closed 1 month ago

fanglin1 commented 1 month ago

In the TrnData class of the DataHandler.py file, the self.negSampling() function should be added during initialization, otherwise the generated negative samples may all be zeros.

ZzMeei commented 1 month ago

Hi,

Thanks for your questions. At the beginning of the training process, self.negSampling() function will be called to generate negative samples https://github.com/HKUDS/AdaGCL/blob/6359f1ed568bf3f99d6351848b9a0c7391c1c93f/Main.py#L78

fanglin1 commented 1 month ago

I'm sorry for my ill-considered question, as I didn't entirely check your code. Besides, thank you for your detailed reply and wish you a happy life!