Note: The following is mainly a collection of advice and resources I maintain for my graduate students.

Guidance on How to Write A Research Paper or Thesis

General Structure of a Paper

A guaranteed way to fail with a paper is to follow the paper style "In this paper, I tell you some selected thoughts about something".
Researchers read research papers not because they are interested in what you think (unless you are a famous visionary), but because your validated findings help them make progress in their work.
This is why the following structure should be followed:

1 Introduction: What is the problem? (Focus on ONE narrow problem)
Why is it important? (Relevance of the problem in research or economic terms)
How does this problem relate to the general state of the art in the research community?
What is your exact contribution?
2 Methodology: How do you come up with your propoposal? (You should follow a rationale when building any proposal for a new solution)
How will you measure the quality of your proposal (define metrics for the evaluation)
3 My Proposal: In here, describe your solution to the problem. Start with an overview of the general idea. Clearly separate conceptual issues from implementation aspects.
4 Evaluation / Findings: In here, you describe what happens when you apply your proposal to the problem. This should especially contain the results of measuring your proposal using the metrics defined in section 2. Highlight suprising findings in here, but do not discuss their implications. That's in the next section.
5 Discussion: Discuss whether your proposal has solved the problem in a convincing manner, use the data from section 4 to substantiate your argument. Point to strengths and weaknesses, discuss what you have solved and what remains unknown.
6 Conclusion: Summarize what you have shown ("In this paper we have shown that drinking water from the tap reduces energy consumption during ontology engineering by, on average, 50 %. We have proposed a new methodoloy for .... ")

All sections except section 1 and 6 should start with a concise summary of what is in that section, so that a reader can skip sections easily.
"In this section, we propose a methodology for the evaluation of the impact of drinking water from the tap on ontology engineering, especially with regard to power consumption."
I hope that helps you get papers accepted!

Research is Taking Place in Communities

Since research is taking place in communities, you must be careful to know and respect the implicit assumptions of the respective community. Otherwise you mark yourself as an outsider and drastically increase the likelihood of rejection.
Just a general guideline:

Don't write lengthy paragraphs about stuff that everybody in the community knows.: E.g., don't write two pages about what the Semantic Web is when writing a paper for ISWC, and don't write a lot about what you think e-business is when targeting an Information Systems conference. You bore reviewers and, unless you are a very experienced senior, will likely write stuff that is wrong or imprecise. A good grasp of an overall research topics develops only over years.
Include short summaries of concepts you are borrowing from other communities.: Stuff that you/your community knows might not be widely known in other communities. For example, when submitting a paper to a library science conference, then you should introduce ontologies in a concise manner. Otherwise, reviewers might not understand your paper and, as a precaution, will often reject your work.

Additional Resources

Alan Bundy: How to Write an Informatics Paper http://homepages.inf.ed.ac.uk/bundy/how-tos/writingGuide.html

Some tips from an older e-mail written by Livia Predoiu: (very much recommended!)
Talk about "How to give a great research talk? How to write a great research paper?" held by Simon Peyton Jones from Microsoft Research in Cambridge.
"How to give a great research talk?" (http://www.wit.at/events/peyton-jones/index.html)
The slides for the presentation can be found here:
Giving a Talk (http://www.wit.at/events/peyton-jones/Giving%20a%20talk.pdf)
Writing a Paper (http://www.wit.at/events/peyton-jones/Writing%20a%20paper.pdf)

They also recorded the presentation on video. You can find the videos of the presentation here:
Video of the Talk: Writing a Paper (http://www.wit.at/events/peyton-jones/writing_a_paper.ram)
Video of the Talk: Giving a Talk (http://www.wit.at/events/peyton-jones/giving_a_talk.ram)
The sound is very poor, unfortunately, but the talk is still understandable. And you need at least Real Player v10.

Some more resources added by Uwe Keller:
- An how-to page on writing Informatics papers (http://homepages.inf.ed.ac.uk/bundy/how-tos/writingGuide.html) by Alan Bundy (http://homepages.inf.ed.ac.uk/bundy).
- The Researcher's Bible (http://homepages.inf.ed.ac.uk/bundy/how-tos/resbible.html) by Alan Bundy, Ben du Boulay, Jim Howe and Gordon Plotkin.
There is also a PDF version (http://www.ifs.tuwien.ac.at/~silvia/research-tips/researchers_bible-1999.pdf) of the document available.

You may also consider to visit David O'Sullivan's course Publishing Your Research - Seven Tips for Effective Writing (http://www.owl.ie/dos.nsf/researchreports?OpenPage) on November 14, 2007

^ TOP

Tools and Resources for Researchers

The following is an unsorted list of links, tools, and data resources which I consider useful for my research:

Graphviz: Open source graph (network) visualization project from AT&T Research
http://www.graphviz.org/
Jena: Semantic Web Framework for Java
http://jena.sourceforge.net/
Protégé: A free, open source ontology editor and knowledge-base framework
http://protege.stanford.edu/
Twinkle is a simple GUI interface that wraps the ARQ SPARQL query engine. It is useful for learning the SPARQL query language and experimenting with Semantic Web data.
http://www.ldodds.com/projects/twinkle/