Scientific Reporting: Mobile Robotics CSC 290 (Fall 2000)
The purpose of a scientific paper is to report results in a way that
they can be completely understood, and if possible, duplicated by
someone with adequate knowledge and equipment. Ideally, to be
publishable it establishes some new fact or hypothesis. Practically,
one usually motivates the questions and shows why the results are
important. Adequate reference to prior work is necessary (since
you don't want to repeat old work or appear ignorant of current work,
especially any work of the people refereeing or reading your paper).
You can get an idea for the general structure of a paper from
looking at some. In certain fields the format is pretty standard, but
presumably there's always room for some innovation or formal
creativityas long as the story gets told.
The sort of things your readers want to know for a paper with a
theoretical and experimental componenent are naturally organized
something like this (some sections may be irrelevant for some reports).
-
Title: snappy, informative.
-
Abstract: a short paragraph telling the punchline, the highpoints,
of the paper.
-
Introduction or Background or Hypothesis:
Short history of the problem you address, what your contribution is
and why it is important. Sometimes the last paragraph
is a little prose outline or reader's guide saying what is in all the upcoming
sections.
- Prior Work. Who has done what in this area (with references)
- The Problem. A formal problem statement, how you are modeling
it with mathematics or algorithms. Try to stay away from
the implementation level here: the idea is to state things in an
abstract way that can be related to other general problems out in the
world. Strive for as general a statement as you can and be as
specific as you can about exactly how you formalize or mathematically
model physical or psychological concepts.
- Implementation. How you translated your abstract problem into
computer code or hardware, what algorithms you used, what
approximations, etc.
- Results. What happened, performance, graphs, images.
- Discussion. Observations about the problem, performance,
methods, what was
easy and hard, what you learned, what the reader should pay attention
to. Can be merged with Conclusions.
- Conclusions and
Recap what you did, why it's important, the main results, and
what conclusions you can draw. Like they contradict someone's
previous idea, or they confirm it, or they extend it. Here you put
the meat of your work. Assume this is all that will get read.
- Future work (sometimes merged with Conclusions):
What should the excited reader run off and try to do
next? What is left undone in your implementation or formalization?
What is the highest-leverage next step?
- Acknowledgements: to referees, colleagues, funding agencies.
- Bibliography
Mobile Robotics Page
This page is maintained by
CB.
Last update: 3.13.98.