3

I have heard that journal papers are more valuable than conference papers generally in other fields. Yet I read that in Computer Science field that conference papers are more valuable.

  • Hence I want to clarify exactly what is true?
  • Also if an extra importance does come to conference papers vs journals why so?

I m in the middle of my research work of AI pattern recognition and I am in the process of deciding to write a journal paper or a conference paper. Already in the field of my research alot of publications exists, yet I am doing a partial novel research, a tiny change to methodology. I want the end user to get an understanding of my data, also I want to get a permanent research job, then I wanted to know will publishing a journal or a conference paper look better in my cv? In the case of explaining my data to end user a journal seems the best option with a greater page count than a conference paper yet in terms of what the internet states the conference papers are better for CS Field, a conference paper would get more value to my cv than a conference paper. When I mean value in aspect

  • to my CV
  • and other aspect to provide the end user a complete understanding of my research
4
  • CS is a fast moving field, and conferences are where you get the greatest and latest news. Traditionally, journals take too long to publish an article and they are used for archiving 'good' ideas only. Also, some CS conferences are ultra selective, and thereby giving them certain prestige. Commented Jul 12, 2021 at 3:37
  • 3
    It is true that some conferences are more prestigious to publish your results in CS than some journals (and that it is the norm in many subfields that effectively all significant research gets submitted to a conference, with a longer journal version sometimes but not always being written up later). However the statement certainly would not remain true if we replace “some” by “all”. Thus, while the statement you are alluding to has a kernel of truth, it is phrased in a misleading way.
    – Dan Romik
    Commented Jul 12, 2021 at 4:10
  • 1
  • Computer Science is a vast field. What you say holds true for some domains like AI and Computer Vision. This may not be true in the case of Theoretical Computer Science (though I am not sure about this).
    – Academic
    Commented Jul 12, 2021 at 6:45

1 Answer 1

0

It might be true for some subfields in CS (like AI, ML, DS, CV, NLP, etc) because if we compare an A* journal with an A* conference:

  1. A* conference does not accept papers just because they are novel but also because they are at the top xx% -according to the judgement of the reviewers, which is most of the time subjective-. For example, if you submit a good paper to a conference with a usual acceptance rate = 15%, but there are 16% of submitted papers better than yours, your paper will be rejected. This means that the accepted papers are outstanding (again according to the judgement of the reviewers). A journal from the other side accepts a paper without a direct comparison with other papers.
  2. As mentioned already in the comments, CS is a fast-moving field and conferences are much faster with the review process than journals.
3
  • I don't think journals and conferences differ regarding point 1. Editors receive many, many more "good" manuscripts than they (can) publish (given a limited number of issues per year). Editors will compare the quality of each submitted manuscript that gets positive reviews to the average quality of the articles that are published; and they make the final decision what to publish or not (rather than the reviewers). The comparison may be more implicit, but it's the same comparison and the same competition. Commented Jul 12, 2021 at 7:41
  • @henning Are you doing this as an EIC?
    – Yacine
    Commented Jul 12, 2021 at 8:06
  • @Younes so because the acceptance rate in journals are higher than in conference papers thats why conference considered as more valueable .Thanks Commented Feb 10, 2022 at 4:16

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