In scientific writing, chapters / sections have a typical role.
Knowing what that is will help you fill these chapters in the expected
- Introduction: motivation.
Finish with an explicit (bullet-point) list of
- What is the problem?
- why is it relevant?
- What are you doing about it?
- Background (optional): describes what is commonly
The point of the optional background chapter is to
present that information which is needed to understand the
rest of the paper, but is already commonly known to those
in the field. E.g. the sort of information you'd find in
any textbook on the subject. Information that is
actually new goes in later chapters.
Background and related work could be one
section/chapter. Typically, background is discussed before
- Related work: (I) showing you know and understand
related work, (II) newness.
In related work you first of all show you understand
work on this subject well. This not only entails
describing what other researchers have done, but
grouping other studies logically together and describing
the various groups in a cohesive whole.
The second point is actually the main goal of related
work: showing that what you're doing is new (and
aligning with previous research).
- Methodology: how and why.
In the methodology section, you compare various
possible methodologies, explain their strengths and
weaknesses and motivate why you choose a particular
- Main part: your actual work.
In the main part you explain your setup, run your
experiments, design your protocol, etc. This could
include any of the following, depending on your subject:
Or, when writing a security analysis:
- Design: in general, more scientifically
- Development: in general, less scientifically
- Setup (of experiments / tests)
- Analysis of results
- Validation: how did you ensure that the results
- Intruder model: what are the intruder's
- Security requirements
- Security analysis: why is the proposed thing
Argue this using (I) intruder model, (II)
security requirements, (III) system
- Discussion (optional): for discussion.
In the optional discussion section, you can discuss
items that did not fit into the rest, should be
mentioned, but don't warrant.
- Future work: next scientific steps.
Future work should look ahead at what scientific
questions can be done based on the work you've
done. It could include some engineering aspects
(refactor, incorporate XYZ subsystem, ...), but that is
not its main point.
- Conclusions: scientific repeat.
Conclusions repeat the key findings... *and* their
impact. Typically, the impact is more important than the
main findings. For example, "we found 37 sites out of 50
were vulnerable" is what you did. To that, you add what
this means: how worrisome is this? what should be
done now that we know this? How can we improve the