44.txt

This complete matlab for neural network

发信人: nohau (踩高跷的鸟), 信区: DataMining
标  题: How to Get a Paper Accepted 
发信站: 南京大学小百合站 (Wed Apr  9 15:44:55 2003)

[I found this on the web somewhere, and didn't want to risk losing it so I sha
melessly created a private copy.] 


Comments by Kent Beck 
I will not talk about a topic area, like my distinguished fellow panelists. I 
will present the process I use as I am writing my papers. You can adapt it for
 your writing process, or you can use it as a check list for evaluating finish
ed papers (if this is starting to sound like patterns, well, fancy that). Much
 of what I will say is "common sense," found in any book about writing. Having
 looked at hundreds of submissions, though, I can state with certainty that mo
st of the authors don't follow this advice. 
Write to the program committee. Never forget that before you can write to the 
vast, eager, and appreciative OOPSLA audience you must first get past the prog
ram committee. Before I begin I fix in my mind a picture of a harried PC membe
r, desk piled with papers. Mine comes to the top. I have maybe thirty seconds 
to grab their interest. Remember that the program committee is made up of expe
rts in the field. Even if your topic is of broad interest to beginners, there 
must still be some spark in it to keep an expert reading to the end. If your t
opic is highly technical, it may not be in an area that they are familiar with
, so it must readably present the novel aspects of the work. 

One startling sentence. Now that you know you are writing to the program commi
ttee, you need to find the one thing you want to say that will catch their int
erest. If you have been working on the world's niftiest program night and day 
for five years, the temptation is to include absolutely everything about it, "
The Foo System In All Its Glory." It'll never work. I know it's painful to ign
ore all those great insights, but find the most interesting thing you have don
e and write it down, "network garbage collection is fast and easy." You want t
he reader's eyes to open wide when they realize what it is you've just said. I
 think some people are reluctant to boil their message down to one startling s
entence because it opens them up to concrete criticism. If you write about the
 Foo System and someone says it isn't neat, you can just reply, "Is so, nyah!"
 If you say network garbage collection is easy, it is a statement that is obje
ctively true or false. You can be proven wrong. Wait! You spent five years pro
ving it was easy. Make your case. 

Argument: problem, solution, defence, related work. Now that you have a startl
ing sentence, your paper must stand as the argument for its validity. You are 
convincing the by-now-intrigued committee member of the truth of your amazing 
statement. Divide your paper into four sections. The first describes the probl
em to be solved. When the PC member is done reading it, they should understand
 why it is a problem, and believe that it is important to solve. The second se
ction describes your problem. You are convincing the PC member that your solut
ion really could solve the problem. This section is sometimes supplemented wit
h a section between the defence and related work which describes implementatio
n details. The third section is your defence of why your solution really solve
s the problem. The PC member reading it should be convinced that the problem i
s actually solved, and that you have thought of all reasonable counter argumen
ts. The final section describes what other people have done in the area. Upon 
reading this section, the PC member should be convinced that what you have don
e is novel. 

Abstract. The abstract is your four sentence summary of the conclusions of you
r paper. Its primary purpose is to get your paper into the A pile. Most PC mem
bers sort their papers in an A pile and a B pile by reading the abstracts. The
 A pile papers get smiling interest, the B pile papers are a chore to be slogg
ed through. By keeping your abstract short and clear, you greatly enhance your
 chances of being in the A pile. I try to have four sentences in my abstract. 
The first states the problem. The second states why the problem is a problem. 
The third is my startling sentence. The fourth states the implication of my st
artling sentence. An abstract for this paper done in this style would be: 

The rejection rate for OOPSLA papers in near 90%. Most papers are rejected not
 because of a lack of good ideas, but because they are poorly structured. Foll
owing four simple steps in writing a paper will dramatically increase your cha
nces of acceptance. If everyone followed these steps, the amount of communicat
ion in the object community would increase, improving the rate of progress.

Well, I'm not sure that's a great abstract, but you get the idea. I always fee
l funny writing an abstract this way. The idea I thought was so wonderful when
 I started writing the paper looks naked and alone sitting there with no suppo
rt. I resist the temptation to argue for my conclusion in the abstract. I thin
k it gives the reader more incentive to carefully read the rest of the paper. 
They want to find you how in the world you could possible say such an outrageo
us thing. There are my four steps to better papers. You can use them sequentia
lly to write papers, or you can use them to evaluate papers you have already w
ritten. 


--
蒹葭苍苍，白露为霜。所谓伊人，在水一方。

溯洄从之，道阻且长；溯游从之，宛在水中央。
※ 来源:．南京大学小百合站 http://bbs.nju.edu.cn [FROM: 192.11.189.114]