Thesis

The following text is only the opinion of MINE research group! Other colleagues may handle this in a different way!

Motivation

The PhD or Master or Diploma or Bachelor (= PMDB) documentation is a final one that summarizes the work the student has performed in the past. Depending on the academic level, it is more scientific or industrial oriented.

The PMDB documentation reflects a work that contains a story, a red threat. Depending on the academic degree to be achieved, it also summarizes the own approach in comparison to other approaches.

For an independent reader who is familiar or even not familiar with the topic, the architecture of the documentation is crucial for his/her understanding. In this respect, the most important intent must be to structure the document in a way that the user understands the things that the candidate has done. A suggestive framework can be found below.

Be aware that the documentation is always the last logical step in your studies, so take care and time to do it (but finally come to an end and do not try to find the grain of sand!).

A good start to perform the PMDB documentation is surely the 'W'-questions like why, who, when, what, etc.

Suggestive Framework (to be adapted)

First of all: Title, your name, work and place
Preface (see comments below)
C1 - Motivation/Introduction: why is your work interesting? Why is it worth to go for this problem? Brief Introduction to the topic! Can you give some kind of recommendation how to read the documentation (for experts/non-experts)?
C2 - Embedding: are there any approaches that go into the same direction than your work? Are there any contributions that reflect the same or similar approach? What is the state of the art in that field? What are challenging approaches? Is your contribution part of a project? If yes, please give a short summary of the project.
C3 - Conditions: explain those things that must be known beforehand, in order to understand your contribution.
C4 - Architecture of your work: how does your contribution work in general (on an abstract level)?
C5 - Implementation Details: here you should explain, how your contribution works in practice. Can you give some comments about the costs, correctness, completeness, data structures etc. and why you have chosen these components?
C6 - Validation and Testing: here, you should prove that your contribution is correct; the contribution should also be explained on - at least - one example. If the contribution is used in practice, you should mention that.
C7 - Future Work: what are the disadvantages and advantages of your system (e.g. in comparison to the current state of the art). What could be improved? what are challenges and visions?
C8 - Summary: summarize the main contents
Appendix: Table of Contents, Table of Figures, Index, Glossary, References, etc.

Recommendations

Always try to write in an objective way and avoid personal (subjective) comments! As an example, sentences like the software test could not be performed because the hardware was so awful and the I could not find a suitable software support should be avoided; a sentence like Because of technical problems, the test could not be performed. should be used instead.
In case you start the documentation with a preface where you want to say thank you to several persons, you probably should avoid the word thank you or thanks to but more try find nice stories/anekdotes in the past. And don't forget those people who supported you throughout your life (parents, friends, etc.). In this respect, try to be concrete and avoid sentences like ...was very nice because he was there whenever I needed him...: this sounds strange and lets this person be someone who seldom gave support and advise.
For large documentations, it is rather more appropriate to select a documentation software like LaTeX or TeX that are - per se - created for larger documentations (than for example Microsoft Word).
The usage of pictures is highly welcome - if it supports the reader in understanding your contribution. Please note that the pictures should be serious and should not represent 'fun', so black-white pictures are generally sufficient. However, coloured pictures must be used in case it is needed.
Concerning the style: there is often the question of using the 'I have shown that'-, 'we have shown that', or a passive ('it was shown that') way to transfer the information/knowledge to the user. It is hard to say, I personally prefer the passive version but it is also possible to use the alternatives.
PMDB-documentations should be started right in time and there should be enough time for this. To be more concrete, I advise to start the documentation from the first day you work on the thesis, writing down some keywords/headwords, and notes. Even the usage of a Wiki-system might be very appropriate, or a web-based system. The number of time to perform the documentation depends on many things, but from my experiences, I believe that a PhD takes 4-6 months, Master/Diploma 2-3 months, and Bachelors 3-4 weeks.
In concern of the number of pages: I am often asked for this (Sir, how many pages do you expect?); it should be clear that the one and only trigger is the topic of the thesis, and depending on the motivation, conditions, etc. the number of sizes change. A work of 50 pages can be better than a work of 200 pages (and vice versa). Some experience values for computer science are
- Bachelor: approx. 50-100 pages
- Master/Diploma approx. 80-150 pages
- PhD 100-300 pages
Writing Strategy:
- start/keep on writing
- after a while, insert conscious writing-breaks in order to get some distance from the topic
- check the reading and find the delta (= what could be improved, what must be inserted, is the style good enough, does the reader understand, etc.)
- if the delta is big enough, then go back to the first point; otherwise finish.
In concern of the language: choose the language you are most familiar in or where you think to be familiar in. English is generally the best choice if you want to broadcast your contribution to others, but external requirements (e.g. industrial requirements) may influence the PhD, Bachelor or Master thesis (Important: please take also into account that the thesis must be reviewed by different persons of different language skills!)
Use the recommendations of your mentor/advisor/supervisor, since he or she has already gone through this process!
It is recommended to read at least 1-2 similar theses in order to get an idea of a thesis. Maybe, you read some parts of a thesis and ask the writer why he wrote it in this way (if necessary).
Plan the last steps of the project probably:
- when to finish (soft and hard deadline)
- where/when to print
- where/when to submit
- is it necessary to have the summary in three official languages (for Luxembourg: english, german, french)?
- what is to do afterwards (e.g. presentation, publication, etc.)
- For PhD: How to publish the thesis at the end? Are there some guidelines?

Evaluation of the contribution

This is a very difficult topic. In my opinion, a Bachelor, Master, and PhD-contribution is

excellent if
- the work is a scientific milestone (PhD, Master) or
- an exceptional work in the industrial environment that may influence future industrial components (Bachelor, Master).
very good if
- all requested parts were done completely and if there are new scientific components that go beyond the intended academic scope and that were brought by the candidate him-/herself (self-contribution; PhD, Master) or
- all requested parts were done completely and if there are new industrial components that may influence future technologies in industry (Bachelor, Master)
good if
- the requested tasks were done completely; the work is good enough to be used for future work (i.e. where others can base on). This is both for PhD, Master, and Bachelor.
average if
- the requested tasks were done (but not even more) with satisfaction. For example, the work is done but there is some need for re-implemenation since the code is not good enough. This is both for PhD, Master, and Bachelor.
low average if
- the requested tasks were not completely done, but the core of the work can be identified. This is both for PhD, Master, and Bachelor.
to be rejected, if
- significant parts were not done and/or the core of the work can not be identified. This is both for PhD, Master, and Bachelor.

Whereas for master and bachelor works, the thesis may be evaluated after a period, the evaluation for a PhD is ongoing.