%%-----------------%%----------------%%
Journal of Interesting Negative Results
%%-----------------%%----------------%%
In our scientific explorations, sooner or later we try out ideas
which, sometimes for quite deep reasons, fail at the experimental
stage. There is a palpable bias towards positive results at journals,
conferences, workshop and other outlets for reporting scientific
work. Negative results, however, can be just as interesting. We often
learn more from failures than from successes.
We are starting JINR, a new on-line Journal of Interesting Negative
Results in Natural Language Processing and Machine Learning. We hope
that it will become a forum for research ideas which appear good but,
upon rigorous examination, can be shown not to lead to good results.
It may also become a place to publish counter-examples to accepted
but sometimes unproven truths.
While we focus on NLP and ML, we will welcome contributions from
other areas of Artificial Intelligence.
Please visit the JINR home at <http://jinr.site.uottawa.ca>.
%%------%%------%%
We invite submission of new work along the lines of our bias towards
scientifically proven negative results. We envisage four categories.
1. A regular paper presents strong intuitions and compelling ideas,
proven wrong in thorough and well-conducted experiments. The
experiments should be described in such a way that they can be
reproduced. The negative results should be explained and justified,
along with the reasons why the idea did not lead to the predicted
results. The lessons learned should be clearly stated.
2. A communication gives counter-examples to accepted conjectures or
to previously published results. We expect the same technical and
presentation quality as in regular papers.
3. A letter to the Editors makes a statement of opinion on issues
specific to the journal. The editorial board will decide how
appropriate it is to publish each letter.
4. Squibs and Discussions include short presentations of issues and
ideas related to negative results in specific areas or domains, the
effect of such results on research, and other ideas of interest to
the journal readership.
%%------%%------%%
Vivi Nastase and Stan Szpakowicz
University of Ottawa
-- Stan Szpakowicz, PhD, Professor 613-562-5800/6687 /~\ The ASCII Ribbon SITE, Computer Science szpak@site.uottawa.ca \ / Campaign Against University of Ottawa www.site.uottawa.ca/~szpak X HTML Email
This archive was generated by hypermail 2b29 : Wed Apr 19 2006 - 23:43:00 MET DST