3 ms·

I know, I know. And the mediocre academics take mediocre graduate students and the whole thing repeats... thank god a lot of the undergraduate students are wise

I know, I know. And the mediocre academics take mediocre graduate students and the whole thing repeats... thank god a lot of the undergraduate students are wise enough to recognize lame professors. The best students have fun making these frauds look stupid in front of the class. BTW students need to be encouraged to act in such ways. I have been fighting my private battle to get evaluators to actually read 1-3 papers of those they evaluate. Usually if someone is working on pointless stuff, you can see it pretty clearly when you read their stuff. Whatever the case, the simple counting technique must stop. BTW some consistent universally accepted impact factor rating might help in computer science. I don't know.

maurycy 19y ago
The Erdos Number is a good attempt.
- jey 19y ago
  Huh? I don't see how having a lower Erdos number indicates that your work is better. The Erdos number has nothing to do with publication quality nor citations.
  - maurycy 19y ago
    It was joke. More seriously, the Erdos number is indeed far better indicator than the number of publications. Number of citations is not as good as it seems, because not all citations are created equal. I believe that, if we are talking about the details here, the best system should be similar to the Page Rank. Citations from important works are more valuable. Awarded papers, like Nobel or Fields Medal, could get the highest rank. The only problem with this system, which I see, is its mix of dynamic and static nature. The paper's rank can be dramatically changed literally overnight. On the other hand, it takes ages to get the Nobel prize. Not that good when your career depends on your rating. However, it should be quite obvious that brilliant work is not appreciated immediately. So, very likely that the problem lies in the educational system, that requires fast and pretty objective, not necessary reliable, method of rating. Furthermore, I don't know whether there's a correlation between educational successes, and research abilities. I bet there is no. Perhaps it does not really make any difference whether a guy got a Nobel, or two.
    - amichail 19y ago
      You can combine PageRank with a prediction market. People would make bets on the (perhaps relative) PageRank of papers at some future date.
    - shiro 19y ago
      "the best system should be similar to the Page Rank." It's been suggested, but I don't know how popular such index is. http://en.wikipedia.org/wiki/Impact_factor#Other_measures_of_impact http://en.wikipedia.org/wiki/Impact_factor#Other_measures_of...
    - anaphoric 19y ago
      Looks approximately correct. http://www-static.cc.gatech.edu/~guofei/CS_ConfRank.htm http://www-static.cc.gatech.edu/~guofei/CS_ConfRank.htm

maurycy 19y ago

The Erdos Number is a good attempt.

jey 19y ago
Huh? I don't see how having a lower Erdos number indicates that your work is better. The Erdos number has nothing to do with publication quality nor citations.
- maurycy 19y ago
  It was joke. More seriously, the Erdos number is indeed far better indicator than the number of publications. Number of citations is not as good as it seems, because not all citations are created equal. I believe that, if we are talking about the details here, the best system should be similar to the Page Rank. Citations from important works are more valuable. Awarded papers, like Nobel or Fields Medal, could get the highest rank. The only problem with this system, which I see, is its mix of dynamic and static nature. The paper's rank can be dramatically changed literally overnight. On the other hand, it takes ages to get the Nobel prize. Not that good when your career depends on your rating. However, it should be quite obvious that brilliant work is not appreciated immediately. So, very likely that the problem lies in the educational system, that requires fast and pretty objective, not necessary reliable, method of rating. Furthermore, I don't know whether there's a correlation between educational successes, and research abilities. I bet there is no. Perhaps it does not really make any difference whether a guy got a Nobel, or two.
  - amichail 19y ago
    You can combine PageRank with a prediction market. People would make bets on the (perhaps relative) PageRank of papers at some future date.
  - shiro 19y ago
    "the best system should be similar to the Page Rank." It's been suggested, but I don't know how popular such index is. http://en.wikipedia.org/wiki/Impact_factor#Other_measures_of_impact http://en.wikipedia.org/wiki/Impact_factor#Other_measures_of...
  - anaphoric 19y ago
    Looks approximately correct. http://www-static.cc.gatech.edu/~guofei/CS_ConfRank.htm http://www-static.cc.gatech.edu/~guofei/CS_ConfRank.htm