search

KaiyangLi1992 / Privacy-Preserving-Social-Network-Embedding

Implementation of Adversarial Privacy Graph Embedding in TensorFlow
19 stars 8 forks source link

Data labels confusion #2

Open xiaotili0123 opened 3 years ago

xiaotili0123 commented 3 years ago

It's very interesting and inspiring work! I have a short question about the data private labels for two dataset. When I visualize the private labels, I found that the index value for Yale is from 0 to 6 ([0,1,2,3,4,5,6]) and for Rochester is from 0 to 2 ([0,1,2]). So, I thought the classes of these two dataset are supposed to be 7 and 3 rather than 6 and 2. I can see the operation in process_attri.py on labels, but I didn't see why. Could you please give more details about that, thank you so much!!

scottshufe commented 3 years ago

I am also confused about this question. I have read the 'readme.txt' file of these datasets, and I found that the missing data is coded '0'. refer to this url

I think the authors may want to eliminate the zero items in the data, but I can't understand the operations in process_attri.py.