wu :: forums - Print Page


    
      
        wu :: forums
        (http://www.ocf.berkeley.edu/~wwu/cgi-bin/yabb/YaBB.cgi)
      

        general >> wanted >> help on writing a paper
        
(Message started by: towr on Jan 19^th, 2004, 8:19am)

Title: help on writing a paper
Post by towr on Jan 19^th, 2004, 8:19am

I need to write a paper on a project I'm doing, but I don't really have an idea of how to start, nor how to continue if by some miracle I ever do get started..
I was wondering if anyone had any helpfull tips or ideas.

Title: Re: help on writing a paper
Post by BNC on Jan 19^th, 2004, 8:58am

When I write papers, I start by deciding on a general structure (e.g., the section headers).

Next I write the "body" of the work (it's easiest for me, cause it desribes what I just did). The summary comes next.

I then go for the literature scan, in order to produce a comprehensive introduction.

The last thing is the abstract.

One important thing is to decide on the journal you plan to publish at -- each may require different emphsis, specific formating, etc. -- it's a shame finishing the work just to realise you have to re-do some of it.

Hope it helps.

Title: Re: help on writing a paper
Post by towr on Jan 19^th, 2004, 9:35am

I don't have to publish it (thank god), so that's not a problem. It's supposed to be for written for my peers, which would be (theoretically ;)) other university students in/near my field.

The main problem is that whenever I try to write anything I'll go sit in front of a computer, and at the end of the day I'll have as much as I had before I started, nothing.. My innate hatred of writing doesn't really help much there either..
And unlike some other papers I've written this one doesn't really offer the option to introduce it by writing about how much I hate writing (it).. (Not in the last place because it needs at least a smidgen of scientific standard)

Title: Re: help on writing a paper
Post by Icarus on Jan 19^th, 2004, 6:21pm

The best I can suggest (it isn't much) is to pretend you are just describing it to someone casually and write out informally what you would say. Then you take that as a basis for your rough draft. Leave the order the same but decide what you ought to include and throw out. For that, I try to put myself in the place of the listener and see what seems to be sensible, what seems confusing, and what appears to be "much ado about nothing". Next pull out a few highpoints and build the summary/abstract around them, restating them.

Now I plow through and try to rewrite the whole thing, ignoring if it doesn't sound good, and just throwing everything I want to say out there.

If I still have the will to continue (I too hate writing, though one would doubt it from the amount I do), I reread and try to tighten up sentences, fix points here, there and yonder. I keep doing this until the nice men in the white coats arrive and give me one too, though their tailor always gets the sleeves too long and puts buckles in the weirdest places... When I get back from my rest, I submit the result, regardless of how I feel about it (which is always that a kindergartner could have done a better job).

Title: Re: help on writing a paper
Post by aero_guy on Jan 21^st, 2004, 1:24pm

I absolutely loathe technical writing but I am currently fixing a paper that just came back from peer review before publication, putting the final touches on another I am about to submit, and roughing together a third (you do a lot of publishing at the end of your dissertation), so I feel qualified to talk about it.

The key to making it bearable and making it good is to tell a story, much like Icarus said. I start by gathering all of the general information I want to convey, this is generally in the form of figures. Next I decide on a logical way to order things so that it comes out as a story that builds on itself. When I have all the figures arranged how I want them I write a skeleton that contains the hard numbers integrated with references to the figures.

Next I "fill in the blanks". I tell the story represented by all the numbers. Conclusions and abstract get done after you see what the rest of the paper looks like and introduction is a seperate animal.

ANYTHING looks more important and more interesting than having to do this. I have been known to decide it was time to clean then gunk off every key on the keyboard when I am stuck doing it. I feel for you. What is it on?

Title: Re: help on writing a paper
Post by towr on Jan 21^st, 2004, 2:44pm

The current working title is
"discovering inference rules from dutch text", it's basicly copying what was done in "discovering inference rules from text" but for dutch instead of english text..
So my instinct on the report is basicly one line
"I did what they did <insert reference>, but for dutch text"
Somehow I have to blow that up to 10 or 50 pages (I hope it's the former, but fear it's the latter amount)

The title probably isn't that informative..
So what does 'discovering inference rules from text' mean, is probably what you're wondering..
It has to do with question answering. The goal here is that given some sort database you can enter full questions, rather then a list of disjoint keywords, and you get your answers.
My part in it is finding a sort of paraphrases for parts of sentences. For instance finding that "X wrote Y" means essentially the same as "Y's author, X"
This supposedly helps in a case where someone asks "who wrote Y" when the only answer in the database says "X, the author of Y, drove off a cliff"

Wether it actually work in a system is anybodies guess.. Most results as far as I'm concerned look like rubish, but my advisors don't seem displeased with it.

One bright spot since I started this topic, I have something vaguely resembling an intro. Which means I actually started, well, almost actually :P

Title: Re: help on writing a paper
Post by John_Gaughan on Jan 22^nd, 2004, 6:02am

How about you start with an introduction to the original topic, then delve into the grammatical differences between the two languages. Then talk about how you had to modify the algorithm to account for these differences. You could also write a section on the similarities and how that helps.

Title: Re: help on writing a paper
Post by John_Gaughan on Jan 22^nd, 2004, 6:04am

Writing really isn't all that difficult, it's just tedious.

Think about it like eating your vegetables when you were young. If you just do it and get it over with, you can grab a beer and relax, rather than sitting there at the table playing with your food when you'd much rather be out playing.

Title: Re: help on writing a paper
Post by towr on Jan 22^nd, 2004, 6:21am

on 01/22/04 at 06:02:05, John_Gaughan wrote:

then delve into the grammatical differences between the two languages.

Then I'd first have to find out what they are ::)

Quote:

Then talk about how you had to modify the algorithm to account for these differences.

It pretty much worked as is, it's just following a statistical recipy.

I don't really have the idea that I actually did much.. And that makes it difficult for me to report as though I have..

Title: Re: help on writing a paper
Post by John_Gaughan on Jan 22^nd, 2004, 6:28am

on 01/22/04 at 06:21:22, towr wrote:

I don't really have the idea that I actually did much.. And that makes it difficult for me to report as though I have..

Be honest then. Maybe the boss doesn't want it but oh well. For me it's a moral issue. That's why I would never go far as a businessman, lawyer, or politician. Math and science is about truth, or at least what we accept as truth because we cannot prove the obvious, e.g. an axiom.

Anyway, I have a feeling that your paper will be about one page long :-)

Title: Re: help on writing a paper
Post by towr on Jan 22^nd, 2004, 6:43am

on 01/22/04 at 06:28:01, John_Gaughan wrote:

Writing really isn't all that difficult, it's just tedious.

It's both. Actually, it's more the former than the latter even (I don't mind it too much if I'm well on my way). If it was just writing anything, than sure, it might only be tedious.. But I don't think they'll accept "I am a fish" written 300 times. Besides I'm not Arnold J Rimmer..

Quote:

Think about it like eating your vegetables when you were young.

Hmm.. yes.. When your body has a sufficiently high antipathy against brussel sprouts that you almost start vomitting trying to choke one down, I think it's good sense not to try that again.. And I never have..

Quote:

Be honest then. Maybe the boss doesn't want it but oh well.

I don't think I will have much of a chance of ever graduating from university if this project doesn't get at least a passing mark. There isn't really enough time to do another one.. So I'm pretty much stuck with it..

Quote:

Anyway, I have a feeling that your paper will be about one page long :-)

well, there's the title page, a reference page, a table of content page, and maybe even some content, so that's at leats 4 ;)

Title: Re: help on writing a paper
Post by usdragonfly on Jan 25^th, 2004, 10:17am

since you hate writing, type it out as you think of what you want to say! then once you have it all put together and just how you want it write down what you have typed!