I thought that this simple question has been already asked here on Academia, but it seems to that it did not happen (or, I was not able to find it). So, I will ask it.

You have accomplished some good results in your scientific project, and you want to publish a paper about it. How do you decide you would like it to be published into some conference proceedings, or as a journal article, or as a book chapter?

I am considering that all these three options envisage peer review, and are all top-A destination for your paper.

I work on computer science so any particular advice and opinion in this area are strongly appreciated.

  • I think IEEE transactions worth more than other materials. @DavideChicco.it
    – SAH
    Dec 4, 2014 at 22:45
  • 8
    Depending on the subfield of computer science, you may be able to publish the paper both in a conference proceedings and (perhaps after suitable expansion) in a journal or book.
    – JeffE
    Dec 4, 2014 at 22:51
  • I support what @JeffE says if it's a least a bit the standard in your subfield.
    – yo'
    Dec 4, 2014 at 23:19
  • Yes, what @JeffE said. Conferences are good for quickly advertising your results, journals are good for getting proper feedback from reviewers so that you can prepare the final, polished version of your work. Dec 5, 2014 at 0:05

1 Answer 1


As one of your fellow computer scientists, I often think of it this way:

  • Start with a conference paper: the page limits mean you'll likely only be able to present the core elements of the work, but it's a fast route to feedback, publication, and beginning to get your results noticed by the community.
  • Next, upgrade to a journal paper: this is where you put out the nice full version of the work, with all of the details and supplementary information. Generally a bad idea to do before you've got a version in print somewhere else, because many computer science journals are dreadfully slow (I think this might be because we've got conferences for a fast option, so journals feel less pressure than in other fields).
  • Use book chapters only for reviews and similar. The problem with book chapters is that they're generally much more difficult for people to find. Most conferences will either be open access or will go into an archive like IEEE Xplore or ACM DL that most institutions have access to. Books, not so much, and it seems to make a quantifiable difference.

There is also the option of workshops, which you did not mention. These, I think, are a great place for preliminary work or for work that you want to discuss with a very particular community.

  • 2
    That's a very interesting and true point about book chapters. Is there any particular benefit to book chapters? If not, why publish in them at all?
    – ff524
    Dec 5, 2014 at 2:17
  • 4
    @ff524 Honestly, I'm not sure. I've written a number essentially because people that I know are putting together a collection and it's a nice opportunity to get my own thoughts in order writing a review paper. The only one of these with any significant number of citations, however, is the one that's also got a preprint on arXiv. I think book chapters is a publication model that hasn't figured out how to live in the internet age yet, and I don't know what their future will be...
    – jakebeal
    Dec 5, 2014 at 2:46
  • The problem with book chapters is that they're generally much more difficult for people to find — It is? Google can still find the PDF on the author's web page, can't it?
    – JeffE
    Dec 6, 2014 at 1:29
  • @JeffE Assuming you feel comfortable with civil disobedience against your copyright agreements...
    – jakebeal
    Dec 6, 2014 at 2:39
  • 2
    @DavideChicco.it Nope. I mean, you could probably technically argue the point, but nobody that I know in CS would agree with you. Especially because the only 'book' about conference proceedings these days the is the "booktitle" line in the BibTex entry.
    – jakebeal
    Dec 8, 2014 at 4:45

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .