12

Sometimes, I have been noticing some postdoc open position announcements that, besides the usual prerequisites (PhD, an excellent CV, programming skills, etc.), asked the candidates to have publications in the most important conferences in the field.

For example:

[...] Candidates need to hold an earned Ph.D. (or be near completion) in computer science or in a related field with good programming skills in C/C++/Java/Matlab. Successful candidates are expected to be self-motivated and have a good publication record (at least have one paper accepted by premier conferences or journals like KDD, NIPS, ICML, IJCAI, AAAI, SIGMOD, VLDB, ICDE, TKDE, TKDD, etc.) and good command of English. [...]

I was quite surprised to read announcements like this (if one made an excellent research, who cares where it has been published?).

Are cases like these rare or quite common? To get hired for an academic position (like postdoc for example), is it so important to publish in top-level conferences?

Is the conference fame considered more important than the paper scientific relevance for a hiring process?

8

I can't comment on whether the practice of explicitly pointing out that publishing in top conferences makes a stronger record is common, although I suspect it is taken for granted rather than spelled out. But, from the perspective of a hiring committee, being published in a top conference generally indicates that

  1. The research is excellent.
  2. The work has been peer reviewed by other top researchers in the field.

Both of these indicators will give a hiring committee a better feel for a candidate's potential for future top quality work.

By publishing in a lesser conference or journal and without those indicators, a closer reading of the publication would be necessary to make a better judgement.

To comment on your question about whether the conference is more important than the work: for most people, I would assume (and hope) the work itself is more important. But the reality is that hiring committees need to differentiate candidates, and judging by conference/journal quality is one way to accomplish that goal. Additionally, trying to read all candidates' papers, and then trying to judge the worth of the research without necessarily being an expert in the sub-fields themselves, would probably be futile at worst, and frustrating at best.

Finally, strong letters in support of the candidate are probably almost as important as the research itself, especially if the letters point in particular to example publications that demonstrate outstanding work.

|improve this answer|||||
  • 2
    On point of minor disagreement. A close reading of the paper is necessary to make an informed judgment of its quality no matter where it's published. But people are more likely to read the paper if it's published in a trusted venue. – JeffE Mar 6 '13 at 20:00
4

I believe it is not rare, at least when considering the DBWorld calls. CS is oriented towards conferences and people value top conferences more than anything else. Although I am just young researcher, I may say people are going crazy for publishing in top conferences in CS. I saw some professors target only and only top conferences.

I do not know whether this is a good thing to the field and I wish to hear from experienced researcher in the field how things was before 10 or 15 years.

|improve this answer|||||
  • 3
    I saw some professors target only top conferences. — [raises hand] – JeffE Mar 6 '13 at 19:57

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.