Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

Technical White paper: How to write one [closed]

Folks,

What is the best way to go about researching and presenting for a technical whitepaper? I dont mean the format, overview, sections and such stuff.

I've never written one - and I wonder if a white paper needs to be very very generic (conceptual) or specific (for instance favouring a particular tool/methodology)

And if your answer favours the generic approach, I'd like to know how one can research for that. Is it better to focus on a smaller use-case scenario, start small, use a particular tool/method, gain good understanding and then research more and develop a wide-angle view on the subject?

like image 912
Critical Skill Avatar asked Feb 02 '09 09:02

Critical Skill


People also ask

How do you write a white paper abstract?

Abstract -- A one-paragraph description of what the paper is about. Do not state the conclusion here; simply tell the reader what the purpose of the paper is. Customers frequently read only the abstract and conclusion of white papers, so provide material that gives them a good reason to read the details.


1 Answers

Yes, try to read other technical white papers. But don't just read any white paper. Read the better ones. You can usually determined which is the "better" one by checking how many times the paper has been cited (One web site I go to for this is cite seer and google scholar). Some general guidelines would be:

  1. Try to be straight to the point, don't beat around the bush.
  2. Use your acronyms consistently.
  3. Take the opportunity to state the weaknesses of previous methods, as it kinda shows you have put in effort to review/survey other's methods.
  4. A technical paper needs to be very specific. State exactly how your method works, state exactly how you conduct the experiments (so that others can replicate your experiments), state exactly your findings (lots of graphs would be nice), and finally, conclude it in 40-60 words or so.
  5. Emphasis on stuff that are new (Stuff that you are proposing) and less on stuff that are old (That would be your background). Make the distinction clear.
  6. Generally, you don't include your source code in your paper. If you must, published in a web page together with links to your paper.

P/S: My advice is a bit biased towards academic paper. But I think it should apply in your case.

like image 126
3 revs Avatar answered Oct 23 '22 07:10

3 revs