Loading ...
Sorry, an error occurred while loading the content.

Re: GEOSTATS: R. Shurtz and Paper or electronic version of SIC ?

Expand Messages
  • Gregoire Dubois
    ... I can only agree with such mail. We too often learn from succes stories and those who wish to try something unpublished take the risk to repeat a work
    Message 1 of 3 , Aug 22, 1997
    • 0 Attachment
      Mohammad J Abedini wrote:
      >
      >
      > Now look at the traditional paradigm of thinking to accept paper for
      > publication. Almost all papers were talking about their success. If one
      > tries to elaborate on his failure then the chance for publication is very
      > low.
      >
      > In order to have progress in science, we need to have so called "clash of
      > opinion". Writing papers or books or .. to recieve tenure or whatever you
      > name it will get us nowhere.
      >
      > As a reader of published articles in SIC project, what I am really
      > expecting to see is some sort of elaboration on failure and the reason for
      > why this approach fail and that approach give rise to promising results.
      > As we all have access to the original dataset, the beauty of the cited
      > project could be that the claim behind the papers be reproducable by not
      > only writers of articles but also the readers of them.
      >
      > Hope this short comments initiate more fruitful discussion on this
      > important topic.


      I can only agree with such mail.

      We too often learn from succes stories and those who wish to try something "unpublished" take
      the risk to repeat a work done by others and to fail for the same reasons.

      It is certainly a waste of time and energy if such risk is not explained and shown to others.

      When we follow the main tracks indicated by published papers (or a PhD supervisor....), we
      restrict our vision of the problems by not knowing where other secondary roads are leading.

      This can be illustrated by the publications I have collected about rainfall interpolation in
      the main journals for hydrogeology and meteorology:

      various methods have been initially compared on very few datasets. It has been shown
      that geostats were interesting to use (when applied on averaged data which is not a big surprise).
      The amount of papers involving only geostatistics has then dramatically increased
      so that newcomers to such field had a complete biased view on how to approach the problem of rainfall
      interpolation. Geostatistics have been even blindly applied without taking into account basic
      hypothesis like the stationnarity. How often would have an Inverse Distance Weighted interpolation
      method been a more realistic way to interpolate the data ? I strongly believe this would have shown
      big surprises.

      This might be a caricature but I believe it is not too far from reality.

      The existence of AI-GEOSTATS itself is due to the "fashions" and trends generated by the too
      selective approach of international journals.

      I hope such mailing list can justify in a better way the reason of being of existing trends.

      For a specific and simple problem, SIC should be a way to define clearly what are possible
      roads and where these roads are leading instead of telling the reader which road he must take.

      I have been looking these last months for potential publishers of SIC.
      If few international journals (I won't give names) have shown a serious
      interest in it, nobody has really understood that it is not the paper written by
      the participant which is the main point. As a consequence, I first have received
      offers to have "the best papers" selected by an external committee and to have
      these papers published in a journal.

      It is of course nonsense and not at all the purpose of SIC.

      The more papers we will have, the more interesting will become each paper because we
      will have the chance to compare various methods applied on the same dataset.

      Therefore, our condition for publishing SIC is to have all contributions published,
      regardless of the results obtained (unless of course the method used is too poorly described).

      The editorial board of SIC will then have to compare the methods used and analyse the results
      in order to get more general conclusions. Such work makes only sense if the reader can judge
      himself by looking at the source and make his own conclusions.


      Have all a nice week-end,

      best regards


      Gregoire



      <gregoire.dubois@...>
      <Moderator of the ai-geostats mailing list>
      <http://java.ei.jrc.it/rem/gregoire/>
      --
      *To post a message to the list, send it to ai-geostats@....
      *As a general service to list users, please remember to post a summary
      of any useful responses to your questions.
      *To unsubscribe, send email to majordomo@... with no subject and
      "unsubscribe ai-geostats" in the message body.
      DO NOT SEND Subscribe/Unsubscribe requests to the list!
    Your message has been successfully submitted and would be delivered to recipients shortly.