Recently, the application of Graph Neural Networks (GNNs) has become more and more extensive, and it has in many fields without topology,such as NLP. In several experiments, I have not got the STOA by GNNs. Unfortunately, I also don't have some theory or more experiments to support it.

How do I analyze it?How to prove the information introduced by the topology is more effective than noise?

  • $\begingroup$ What is STOA? Do you mean Graph Neural Networks by GNNs? $\endgroup$
    – Nachiket
    Sep 14 at 6:56
  • $\begingroup$ Yeap, any Graph Neural Networks. For example, graph attention networks for text Classification. Obviously, it does not work well on a specific data set. Therefore, I want to distinguish what causes it? Is this data set not suitable for introducing topology or does the GNNs model need to be improved? $\endgroup$ Sep 14 at 12:06
  • $\begingroup$ What is STOA? Doesn't the graph provide a topology to your problem? $\endgroup$
    – nicoguaro
    Sep 16 at 1:26
  • 1
    $\begingroup$ @nicoguaro I have seen STOA (or more often SOTA) used to abbreviate State Of The Art. $\endgroup$
    – Tyberius
    13 hours ago
  • $\begingroup$ @Tyberius, that does not match that much in context. But who knows... $\endgroup$
    – nicoguaro
    8 hours ago

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.