From: dnk@OMIT.cs.mu.OZ.AU   
      
   In <41f4352c$1@news.unimelb.edu.au> steve_giovatto@hotmail.com (Steve   
   Giovatto) writes:   
      
   >Hello guys,   
      
   >I am a student doing a Ph.D. in Computer Science, and only recently   
   >submitted my first paper (to a conference). I was wondering whether   
   >any of you can give me some advice: Assuming I want to pursue a career   
   >in the field (say, become an academic), how important are conference   
   >papers vs journal papers? Obviously journal papers are more   
   >'valuable', but by how much? It seems when I check scientist's web   
   >sites, they have lots of conference papers and a few journal papers.   
      
   >Is this a silly question..?   
      
   No, not at all.   
      
   >I will greatly appreciate any advice you can give.   
      
   Journal papers are usually expected to be comprehensive, archival   
   quality expositions of a reasonably complete body of work on a   
   theme.  Conference papers range across a spectrum from preliminary,   
   speculative ones through reports of work in progress to more solid   
   and complete ones.  Conference papers are usually quite limited in   
   length, and have a reasonably short elapsed time from submission to   
   presentation/publication.  Conference reviewing is highly variable   
   in quality, some CS conferences set quite high standards and have a   
   low acceptance rate, others don't.  Even minor workshops can now   
   usually get their proceedings published in book form whether or not   
   they have significant quality content.  Journals typically have a   
   much more careful and leisurely review cycle and their articles are   
   substantially longer, so journal publications can take literally   
   years from first submission to final publication.  The amount of   
   work involved in preparing and revising a good journal article is   
   of course much greater.   
      
   You should probably think of submitting a conference paper as an   
   opportunity to communicate quickly with your research community   
   about what you are doing, have your work assesssed by others, make   
   contacts, learn more about how to write a good paper, and also of   
   course to facilitate your participation in the conference itself.   
   A journal paper is more a labour of love that requires time, care,   
   solid ideas and results, a good understanding of the research   
   field and related work, and patience to see through the process.   
   As you noted, good researchers tend to publish about 3-5 times as   
   many conference papers as journal papers, and you'll often see   
   that a journal paper is preceded by several conference papers on   
   related topics in which the research theme was developed.  Noone   
   really expects CS Ph.D. students to publish journal articles prior   
   to completing their degrees, but doing so is expected during your   
   post-doctoral activity if you follow the academic road.   
      
   Research institutions and research funding bodies have various   
   formulae for scoring publications to measure research productivity.   
   Journal and book publications are worth significantly more than   
   conference papers, but often all refereed conferences papers (and   
   papers at workshops at conferences) count the same, irrespective of   
   the quality of the conference.  This unfortunately gives some   
   researchers an incentive to turn a few ideas into lots of published   
   papers, differing only in a few details, scattered around various   
   different conferences.  And one sometimes sees people claiming what   
   is essentially a conference or workshop paper as a book chapter   
   because of the higher score.  Poor researchers tend to have   
   publication lists dominated by conference papers at "easy"   
   conferences, often with titles that are suggestive of "recycling".   
      
   Each area of CS tends to have different unwritten rules about what   
   is reasonable publishing behaviour, so I suggest you try to get   
   advice from your supervisor or some other senior researcher in your   
   area about what the relevant conventions are, how your publication   
   record might be assessed by prospective academic evaluators, and   
   what conferences and journals to target.  If your Ph.D. work has   
   produced some solid outcomes, you should be aiming to produce   
   several conference publications during its progress, and then to   
   consolidate it as one or more journal articles once you complete.   
   Unless you're particularly brilliant, lucky or productive, you're   
   unlikely to be able to produce a paper of journal quality until   
   after you've had more experience in writing papers.  An exception   
   to this would be where you publish jointly with more experienced   
   authors because the work is part of a larger collaboration, but   
   that's not so common with Ph.D. work in CS.  If your supervisor   
   has the time to really collaborate with you on paper preparation   
   you might achieve a journal quality paper (but if it's your work   
   you should be the first author!), however it's usually easier to   
   distill your Ph.D. into journal papers after you've finished it.   
      
   Hope this helps,   
   David   
      
   >My email is: steve_giovatto @ hotmail . com   
      
   >Sincerely,   
   >Steve   
      
   [ comp.ai is moderated.  To submit, just post and be patient, or if ]   
   [ that fails mail your article to , and ]   
   [ ask your news administrator to fix the problems with your system. ]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)