Conference Rankings

. Thursday, August 07, 2008
  • Agregar a Technorati
  • Agregar a Del.icio.us
  • Agregar a DiggIt!
  • Agregar a Yahoo!
  • Agregar a Google
  • Agregar a Meneame
  • Agregar a Furl
  • Agregar a Reddit
  • Agregar a Magnolia
  • Agregar a Blinklist
  • Agregar a Blogmarks

One important process in Science is the publication of the results of research. Sometimes, happens that you have written an excelent paper (not to me, I use to work with a deadline in mind in order to work under pressure) but you don't know a good journal/conference to publish the results. JCR listings makes easier some decisions about publishing in journals but there isn't an equivalent indicator for conferences.

One of my Thesis advisors, Ana Iglesias, gave me some good advices in order to choose relevant conferences. "Translating" those advices into rules, a good conference is the one that
Do you use any other ranking based criteria in order to choose a good conference?

According to these rules, the conferences I've attended recently or I will attend soon, are:
  • CIKM: A in Core (OK), top 35% in Citeseer (NO), 0.90 in CS (OK) => OK
  • IDEAL: C in Core (NO), top 96.72% in Citeseer (NO) => NO
  • ECDM: => NO
And the conferences I have sent some papers:
  • ICDM: A+ in Core (OK), top 59.86% in Citeseer (OK), 0.73 in CS (NO) => OK
  • ICDIM: =>NO
Most of the results make sense, CIKM and ICDM are really interesing conferences, IDEAL is an interesting forum as combines information processing, data mining with more applied topics as bioinformatics and financial engineering, but is not a really cutting-edge conference. ECDM is not a good conference and it is logical that it doesn't appear in those rankings. The only conference where I disagree with the previous "rules" is ICDIM. I think ICDIM is a good conference but, maybe due to its short life, it doesn't appear on the rankings.

According to these rankings, the most interesting conferences on Data Mining and Machine Learning are:
  • Citeseer: ICML, IJCAI, KDD, AAAI, NIPS
  • CS Ranking: KDD, CIKM, AAAI, NIPS, IJCAI
  • CORE: It's not a real ranking, but it gives A+ to all the best conferences from citeseer and CS Ranking.

12 comments:

Jose Maria Gomez Hidalgo said...

What is good about the rankings you mention is that the criteria are (somehow) public. But for instance, the promotors of the "CS Conference Ranking" are unkown. You can find many other rankings by Googling its very name.

I can only trust a ranking if the authors are reputable, an important part of the academia agrees on its quality, and the criteria are known and sensible.

JoSeK said...

You're totally right :) I think "CS Conference Ranking" is quite a good ranking as the conferences seems a logical order and you "sense" there is some study behind the data, but it would be neccesary to know about the machinery behind the ranking in order to trust it for more serious questions.

L. Venkata Subramaniam said...

Dont you think there are too many conferences these days in this area. I cant think of any above average paper not getting accepted after one round through even the best confs. If you start with the confs with Jan submissions, you will very likely find that it will get accepted by Dec after running through 2-3 confs. I think we need to raise the bar further and reduce the number of "good" conferences.

JoSeK said...

I don't think there are too many very good conferences in Machine Learning. If you choose the top 10% of the conferences using citeseer ranking, only a few are related to Machine Learning (COLT, AAAI, KDD, IJCAI and ICML).

And, from my experience, it is not easy to get a paper accepted in any of that conferences.

I agree with you in the sense that if you don't care about publishing in a high quality conference, it results easy to publish a certain paper in any of the ML related conferences (there are a lot of medium-low quality ML conferences).

As I previously say, this is my experience, and I know I've a lot of limitations as researcher, the first one is that I like to work in many different areas and I prefer applied research that base research, and I'm not a native English speaker. All these things make more difficult to write an excellent paper to be published in a great ML conference.

L. Venkata Subramaniam said...

No i am not talking of second tier conferences. Among just the top ones like COLT, AAAI, IJCAI, ACL, COLING, KDD, ICDE, SDM, CIKM, SIGMOD, ICML, VLDB if you start with ACL in Jan by the time the full cycle completes your paper would get accepted in one of among this long list.

No other field has this many conferences and in no other field is it easier to get a paper published.

JoSeK said...

If you're talking about the entire AI field, then I agree, but for me ACL and COLING are clearly not ML conferences, and ICDE, CIKM, SIGMOD and VLDB are conferences where some ML papers may have some relevance, but are not real ML conferences.

If you're working on ML applied to a linguistics problem, then you have a lot of possibilities, as your paper can be accepted in a more general AI conference, or you can send it to a ML conference or even a linguistics conference, but that's not the typical case.

Another point is that, even having a lot of conferences, if your paper is rejected with a bad evaluation in any of that conferences, it's almost imposible to get it accepted in another one conference if you don't work hard on the revisor's comments.

As a conclusion for me, all the Computer Science fields have a lot of conferences, as we are typically conference-guided guys. Another fields like Astrophysics or Maths, are journal guided fields, and have much less conferences. But one thing is the existence of a lot of conference and another is that is can be easy to get a poor paper accepted in a really good conference.

Anonymous said...

You may want to know that if you send email to the contact email list on the ranking site you mentioned, you will get answers. I have exchanged about 20 emails with that address these four days. After some back-and-forth, this people eventually claimed that he was Prof. Hamid R. Arabnia, the organizer of worldcomp, a multi-conference held annually in Las Vegas.

I posted part of the story on www.rankingexpose.com today, you may want to check it.

Anonymous said...

contact@cs-conference-ranking.org has been operating as a "restricted" listserv. Many of my colleagues are members of that list. In fact, any one who wishes to provide his/her services will automatically become a member of that listserv. If you are chairing a conference and are volunteering to help, then you'd be asked to help with events that you are not a member of. As expected, the most knowledgeable people (for ranking and evaluation purposes) are those who have been adminitrating conferences (either, as chair and/or steering committee member) and/or acted as editors of journals.
I cannot help feeling that The "Anonymous" who posted the name of a chair of a conference on April 2009 must be a disgrunted person.

Many chairs of conferences would have to reject many submissions (ie, upsetting authors who are particularly close or at the middle of their promotion/tenure considerations. Unfortunately, many papers submitted by a few such desperate people tend to be obvious cases of "self-plagiarism". One would have to assume that anyone who posts such bashes about others, must have been caught red-handed by the conference administrators/referees/chairs.

Having said the above, I agree with both Jose and JoSek's posts.
Cheers!

Julien said...

The question of "Are there too many conferences these days?" has both answers- YES and NO.

True that there are quite a few conferences these days but the "top-tier" conference are still quite limited. And its still quite hard to get your paper published in any of those conferences, particularly if you are an "outsider" like myself.

I have seen ordinary papers getting accepted in "good" conferences because the authors are from Stanford, MIT, uTexax, etc while some (what I consider) relatively good papers being published in an otherwise not so highly ranked conference. Presumably, either the authors got rejected in the top-tier conferences or just chose to publish in a 2nd-tier conference.

Further, conferences like ICML are more and more going towards more theoretical papers with lots of mathematics even if the results may not be that good or prove so useful (practically). In this case, the "other conferences" do fill the gap and given that surely the ML and AI community is increasing day by day, not everyone is able to publish in these "1st tier conferences". Sometimes, we do also have restrictions on funding for example to travel to U.S, Australia, etc (I live in Germany).

Coming back to the question of ranking, I think experience in the field and general concusses in the community is still the best way to rank a conference.

Our team has had a select conferences where we publish like ICML, ECML, NIPS, ICDM, CIKM, ECIR and of late we tried ICMLA for papers with more practical part. Personally, I like this list (http://sites.google.com/site/fawadsyed/fun-stuff) since my area of interest corresponds to his and he has a quite well maintained list.

I tend to agree with his ranking in general. I am not sure if he/she has a scientific basis for it but it seems coherent with the reputation of the conferences.

Anonymous said...

I believe all the conferences should have a blind review process. I mean why most of them are NOT blind review??

Anonymous said...

Hello. Facebook takes a [url=http://www.onlineroulette.gd]slots[/url] stick on 888 casino disburse: Facebook is expanding its efforts to put forward real-money gaming to millions of British users after announcing a prepare with the online gambling companions 888 Holdings.And Bye.

Anonymous said...

Its such as you learn my mind! You appear to grasp a lot about
this, like you wrote the guide in it or something.
I think that you simply can do with some p.c. to force the message home a
bit, but instead of that, this is wonderful blog. An excellent
read. I will definitely be back.

Take a look at my weblog - avoid hives