Open YuWangyin opened 8 months ago
I do not see any code on how you generated the train_list.npy for your datasets. Does this file record all user_id and item_id with interaction records? Or should we only retain data that has been filtered by 5-core?
I am curious, too. It would be great if the authors shared their preprocessing code.
+1
I do not see any code on how you generated the train_list.npy for your datasets. Does this file record all user_id and item_id with interaction records? Or should we only retain data that has been filtered by 5-core?