Open inst15 opened 11 months ago
In fact, you indeed only need to embed only 'realData.csv' in my work. FakeData is actually just links that do not exist in 'realData.csv'. When we embed only 'realData.csv', it not only represents the links that truly exist in 'realData.csv', but also represents the edges that do not exist within it. By using fakeData, we can filter out some of the non-existent edges!
Hi xihairanfeng, thanks for your nice work! I have a question when trying to run the code "DataProcessing.py" to obtain embedding of node2vec in four datasets: I noticed that it embeds only "realData.csv." Do I need to process both "real" and "fake" data, or can I choose one of them? Also, how are "realData" and "fakeData" represented in your paper? I would greatly appreciate your response!