Home · Topics · Link Farm

Thesis Organization

This page contains a tips on how to organize your thesis. It has tips on how to structure your thesis (naming and order of chapters), and what to put in it.

Navigation

Thesis Organization

Organization is the easier aspect because most Computer Science theses follow a fixed format. This format is an expanded version of the virtually standard form for CS papers. If you deviate from this format without good reason, you will almost certainly confuse your readers, or have them think that your work is not worth reading. In detail, this format consists of:

  1. Introduction
  2. Related Work
  3. Method and Implementation
  4. Evaluation
  5. Summary and Conclusion
  6. Further Work
  7. Appendices
  8. References
  9. Glossary
  10. Index

Some slight reordering is possible. For example, you can put the section on Related Work after the evaluation, or you could actually conclude your work with the Conclusions.

This ordering is standard because it allows the reader to build up expectations as they read the thesis and to check their expectations against what they read. Each section has a particular function, and we will go through every section and describe its function.

Introduction

As the name implies, the purpose of the introduction is to introduce your readers to the problem that you have solved with your work. After reading the section, the reader should have a clear idea what exactly the problem is and what you did about it. To this end, it should at least contain a problem statement which describes the problem that you have solved, and a list of contributions that you are making to advance the state of the art.

It does not hurt if you identify those two sections explicitly, for example by writing things like “We are considering the following problem...”, or “Our contributions are...”.

Related Work

The purpose of this section is to put your work into context. No researcher works in a vacuum, and there are always people working in the same area or in related areas, and they may be using the same or related methods in solving their problems. Your work is useless if it essentially duplicates somebody else's work, so you need to show the reader that you know who else is working in related areas, which methods they use, and that your method is different.

Method and Implementation

This should be one of the two large sections of your thesis (the other being the Evaluation). The purpose of this section is to describe what you have done and how you have done it.

That said, writing this section requires a balancing act. On the one hand, you should describe any method in sufficient detail so that someone with a knowledge of the state of the art in your field can duplicate your results. On the other hand, you should not describe much of anything else. This goes especially for the implementation. No matter how big the temptation, do not include Doxygen or Javadoc documentation for all the classes that you wrote, not even in the Appendices. Doxygen and Javadoc are best read online. This way, the drift between code and documentation is lessened. Do not include static UML class diagrams. But do include architecture diagrams.

To illustrate your method, you should use examples. You can and should use simple examples as long as they illustrate a point. It is best, however, if you can find an example that also appears in the Evaluation, even if you have to simplify it somewhat.

Evaluation

you need to demonstrate that your work actually solves the problem. Unless your work is purely theoretical (in which case you simply prove your solution to be correct), you will need to evaluate your method using a selection of specimen data. The purpose of this section is to demonstrate to the reader that you can solve relevant instances of the problem with your approach.

We can't tell you how to select the specimen data because that depends very much on your work. However, the following guidelines apply. (In the following list, we assume that you are analyzing computer programs, so we will use “programs” for “specimen”.)

Summary and Conclusion

The purpose of this section is to wrap up the preceding sections in a few sentences. It serves to emphasize the points that you made in the earlier sections of the paper by repeating them. The Introduction has the purpose of preparing the reader for the paper that lies ahead of him; the conclusion repeats the main points of the paper that now lies behind him.

Further Work

The purpose of this section is to outline work that you have planned but didn't get to yet. If your evaluation has toy problems instead of real-world analyses, this is the section where you explain how you plan to do a thorough evaluation on real programs.

In research papers—but not so much in theses—this section is also used to stake a claim to a neat idea that belongs naturally in your work but which will take some time to implement and evaluate. The idea is that if you get the idea published, you claim it as yours and discourage potential competitors from working on it, because they would have to acknowledge that you were there first.

Appendices

Appendices are sections that contain material that would be too tedious to put into the main text or that explore interesting sidelines that are irrelevant to the main thread. For example, a long proof of some theorem can go into an appendix (unless of course if it forms the core of your thesis proper).

If you have a section that seems strangely out of place even though it is topical, you should think first of deleting the section outright. If you think you can't (or won't), it probably belongs in an appendix.

Most theses can make do without appendices.

References

This section contains all the literature that was referred to in the main body of the thesis. It will only rarely be necessary to put references at the end of every chapter. The section's formatting and layout is best left to BiBTeX and LaTeX. There is also a section on Citations here.

Glossary

A glossary in a thesis is a small dictionary that explains uncommon terms or terms that are used with unusual meanings in your thesis. Being like a dictionary, a glossary is sorted alphabetically. Most theses can do without a glossary; if you use lots of words in unusual meanings, consider using different words.

Index

According to the OED, “[A]n index is an alphabetical list of names and subjects, together with references to the pages where they are mentioned”. The typography of an index is most easily handled with tools like “makeindex” and presents no practical problems, but the preparation and maintenance of a good index is a lot of work. We don't generally recommend that you undertake it, especially since most theses can do without an index.

Valid XHTML 1.0 Strict Valid CSS!