Basically, my research was not into a particularly new methodology, but we have combined several existing methods to prove a result which has not been seen before under certain circumstances (rebalanced dataset, etc). Would this be an acceptable submission to a conference or journal, or should I go in a different direction and look for something unique?

For reference, this is in the field of machine learning / network security.

  • I thin the devil is in the details: it is hard to give a general answer, but it is very common to publish new applications / proofs. Have you consulted with tour supervisor? – Greg Jun 4 '17 at 12:13
  • Straight forward answer is YES. but, novel method is weighed higher than novel results. – Coder Jun 4 '17 at 12:23
  • @Greg My supervisor has okayed it; however, he's an expert in the Network Security domain and not the machine learning domain (where we want to publish). – Novella Jun 4 '17 at 12:26
  • @Novella Even so I am sure he knows more about your topic than random people from internet. Also, consider costs/risks: what do you loose if you learn a new field, put together a decent paper? In most fields you don't loose much. – Greg Jun 4 '17 at 13:06

The point of a methodology is to be able to produce interesting and significant results. If you've managed to apply some combination of existing methodologies to discover something interesting and new, then that discovery is likely publishable.

What is not clear, however, is whether a community of researchers who are focused on making new methodologies will be interested. Thus, you might not be able to publish in the particular machine learning community that you are interested in. Instead, I would suggest:

  • Look for "applications" journals, conferences, or tracks associated with the machine learning community. Many "tool" communities have these associated: for example, in the general AI world, the core conference AAAI is always co-located with the much smaller but still high-significance IAAI applications meeting.
  • Embrace the application community (it sounds like network security in this case), and publish your results there. They will be less interested in the methodologies and more in what you have done with them.

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.