Yao

> ``` > The code for link prediction needs to be reviewed by the company before it can be uploaded, which takes about 1-2 months. > There is little difference...

> Sorry, I am not super familiar with the pytorch implementation, but I think even if for the first batch, the k-means is not done, it can just use the...

Hello I can't understand this paper. Could you send me the slides, please. [email protected].

I have an another question about the code in the part of recommendation in process of dfs，why candidates ignore the user node？I can understand that implement is about the set of I-U-I in this code？Thank you for your help ------------------ 原始邮件 ------------------ 发件人: "ZYang16"