I wrote a 10 page paper on various facts and anomalies that need to be considered for building an artificial intelligence. The paper is a meta-analysis of various other research from biological sciences, psychology, cognition and artificial intelligence history. The editor of an artificial intelligence journal rejected it, offered some praise and suggested that I split the paper into multiple papers, where each paper presents hypotheses and delivers a specific message. He mentioned:

"This is what we call in academia 'a track record'. If you can deliver one message well in each paper, the collectives of the messages delivered through a track record of papers will add to the more complex message that you intend to deliver."

I assume I'd have to split it into three or more papers as My paper title - Part 1, My paper title - Part 2, My paper title - Part 3.

But if I do that, would it suffice if I mention "Related work" (references to other papers that have attempted similar research) only in the Part 1 paper? And in Part 2 I could just start with a small introduction and then directly delve into the meat of the paper without mentioning any "related work"? Are there any rules or standard practices that allow considering the three papers as one entity, so that the "related work" would be mentioned only in the first paper? I guess the "Conclusion" section would be unique to each paper, so that's not an issue.

  • 1
    The Google Scholar search '"artificial intelligence" intitle:"part 2"' produces 668 results. Why not take a look at some of those papers and see how they handled this situation? Sep 12, 2020 at 12:14
  • The papers are behind a paywall. The non-AI papers that could be openly accessed were papers that didn't have the usual structure. I'll try searching for more.
    – Nav
    Sep 12, 2020 at 12:43
  • 1
    Is the paywall a problem because you're in too much hurry to wait for an inter-library loan, or is it a problem because you don't have an account at a library where you can request inter-library loans? Sep 12, 2020 at 14:32
  • 1
    ... and before concluding that the papers were behind a paywall, did you check the "All n versions" links on Google Scholar? (Although admittedly, sometimes the free-of-charge full-text copies one finds that way are of questionable legality.) Sep 12, 2020 at 14:40


You must log in to answer this question.

Browse other questions tagged .