Tips on projects

This page is for students who are about to start their MSc projects. However, some of the tips could be useful for undergraduate projects as well. I have supervised many students, and often I had to give the same advice. So, this page is for students to read and for me to save time not repeating the same thing again. Also, I assume that you have already read the official handbook for the project module. This page is NOT a substitute for the official guidelines of the module handbook. Rather, this is a complementary collection of some practical tips from my personal experience as a supervisor and a marker presented in a less formal language.


Choosing the topic

This is arguably the most important step in your project, and it is better if you start thinking about the topic long before you have to submit it to the project manager. When choosing the topic, you have to avoid by all means three things:

  1. Topics that are NOT interesting to you
  2. Topics that are ONLY interesting to you (but not interesting to your supervisor, your friends and other people)
  3. Topics that are unfeasible (e.g. flying faster than speed of light or making gold from bronze)

Some students have clear ideas about their interests and have picked the subject during their course or even before. If, however, you find it difficult, then you should know you are not alone. There are many students who are scratching their heads about this. A good question you have to ask yourself is Why are you doing this project? If it is just to get a degree, then ask yourself Why do you need the degree? My advice is try to learn something that will be useful for you in the future. Ask yourself Where do you want to work in the future? The best thing you can do for yourself is to become an expert in the area where you would like to work before you apply for the job, and your project can be the key to it. Imagine how useful it will be if your CV shows to your potential employer that you are an expert in this area already.

After identifying the area your project will be in (e.g. networks, artificial intelligence, web-programming, etc), the next thing is to identify the problem your project will solve. Yes, you need some specific problem. After all, solving problems is what this is all about. However, it does not have to be some "boring problem". You may teach a robot to play football and be more intelligent, or how to classify pancakes. It is all up to you.

Many students think that choosing a subject area, such as Artificial Intelligence or Internet Security, is sufficient. The question they have to answer next is what exactly will they do in this area? Writing another thesis about how important the Internet security is or how exciting is artificial intelligence is useless because many clever people have done this already and written huge books about that. (Are you sure you will do it better?) So, don't try to re-invent the wheel. Use it! Remember that any area of science or technology does not exist for very long if it is not helping to solve some real problems, and we are not teaching you about them just because we like them. In fact, the problem may even come first before you choose the subject area, because the solution will define which areas of computer science you will need to use. One great physicist said:

"How wonderful we have met with a paradox, now we have some hope of making progress" - Niels Bohr.
Top

Writing a proposal

Your proposal is needed for at least two reasons:

  1. It allows you to formulate your initial idea and show it to your potential supervisors (you cannot discuss the project without at least some outline of what you want to do).
  2. It should convince your potential supervisor that you are serious and can work (if you cannot write a proposal, how will you convince them you will be able to write a thesis?)

Your proposal should convince the reader that what you want to do is important (and not only to you), that you have some ideas about how to do it and it is feasible. Thus, ideally, your proposal should not only discuss what you are going to do, but also make some assumptions about how you will do it. This also may include some estimate of what you will need (e.g. library resources, data, software, etc).

When writing a proposal, I advise you to think about the following questions:

  1. Why do you think this problem is important?
  2. Who is it important for?
  3. Who else it might be important for?
  4. Is it a feasible problem?
  5. Has anyone tried to solve this problem already? If yes, what were their results and what will be different/better in your approach?
  6. Can you estimate what do you need to realise the project, and do you have (or will you have) sufficient resources and expertise for it?

Note that what you write in the proposal does not oblige you to do the work the way you have described. It is quite possible that your methods, problem definition and possibly even the subject will change at a later stage. Do not be afraid to make a mistake (you are learning) or express raw ideas (at least you will show you have them). Your supervisor will be there to tell you if you are on a wrong path. The proposal should show that you know where to begin. Without it you cannot start anything.

After discussing your initial ideas, you possibly will have to make corrections or even think of another idea altogether. After the proposal has been accepted, you will need to write a formal proposal. This should also include some time-line describing what will you deliver and when (the milestones).

Top

Resources

You will have to think in advance of what you will need to implement your project. You may require literature, data and software.

I don't know why, but many students think that Internet is the only resource they need to use. Perhaps, this is due to its convenience, but there would not be a need to pay high student fees if Internet was the only resource the university provided to you. Internet is a very useful tool to search for some initial materials, but remember that most of really valuable information is not free (unfortunately). This includes books written by international experts, scientific journals with the latest research and data banks. However, being a student of Middlesex University you are entitled to use these resources. One service that is provided, and not all students seem to be aware of that, is the access to ATHENS - a huge collection of journals and databases that university purchases for you to use. So, go to the library or on their website, read about how to log in and use it.

Another good starting point is scholar.google.com. This is a search engine specialising on scientific journals, books and other peer-reviewed materials (see note on peer-reviewed in the references section).

Top

Project management

This is your project (not your supervisor's). So, it is up to you how you will manage your work. Your supervisor will not remind you about a deadline, because it is your deadline, not theirs. Remember that it is your degree that is at stake, not your supervisor's.

Do not expect that supervisors will spend unlimited time with you. In fact, there is a very limited amount of time available for contact because academics are incredibly busy, and your supervisor is likely to supervise a dozen of other students. So, you have to make the most from the meetings available. Prepare for the meetings in advance. Bring with you what you have done so far or even send in advance by email. Write down questions that you will want to discuss. Write down the comments you receive at the meeting.

When submitting your work for a review or an advice, allow some time for the feedback. Never ask to give the feedback on your report one day before the deadline you have to submit it. Allow at least one week for it (your supervisor may have a dozen of reports like yours on top of the rest of their work).

Top

Writing and formatting the report

The report is a formal document, so do not use informal language. A good reference point about the language and formatting issues is the APA manual (the American Psychological Association style), which is adopted by many journals in different fields, not only psychology.

A well-formatted document can help your project to be properly assessed. There are many rules, but here are some basic tips:

Typically, a report has the following structure (refer to the BIS4992 handbook for more details):

  1. Abstract (About 500 words about the project)
  2. Introduction (Introduce the problem and overview the report)
  3. Literature review (Overview the area, how and who tried to solve the problem before. This can be a part of introduction)
  4. Methods (Describe your methods, solution or the algorithm)
  5. Evaluation / Implementation (Describe how your system works or report the results of the tests and analyse them)
  6. Discussion / Conclusions (Discuss your results, evaluate potential problems and future work)
  7. References
  8. Appendices (You can leave some details here)

My advice is to start by writing the substantial chapters (i.e. the literature review, methods and evaluation) while you are doing the work. Leave the introduction and conclusion for later because often many things can change during your work (even the problem definition).

It does not matter what software you are using to type your report as long as you find it convenient, but make sure you back up your work. It is also a good practise to split large documents across several smaller files. If you are going to work in the industry, then perhaps the popular word processing programs should be sufficient. If you are thinking to continue your career in academia, then consider using the LaTeX system. It has a lot of useful tools, such as BibTeX for managing bibliography and references.

Refer to the APA manual for other formatting and language issues.

Top

References and citations

Students tend to underestimate the importance of proper citations and referencing. Often it is possible to say a lot about the work just by looking at the list of references (and, in fact, this is often what many of us do after reading the abstract and before reading the report itself). In general, it gives a pretty good idea if you understand the basics of how researchers should ground their arguments, and whether you have explored the field deep enough. However, don't include in the list of references every book or Internet page you have read. There are two main reasons why you should cite somebody's work:

  1. Justify your claims and arguments by referencing previous results.
  2. Acknowledge the contribution of other people that is important to your work.

It helps sometimes to think of the references as witnesses in court giving evidence to support your case. In general, any claim you make should be supported by some source. For example, don't just make a statement, such as "every year, more and more people use the Internet". You have to support such a statement by figures and reference the source where you got these firgures from.

Another common mistake students do is referencing mainly Internet pages. The main problem with such citations is that pages on the Internet are often not peer-refereed. For example, I can publish a page claiming that I have invented a way to travel in time, and you can reference my "work" in your report. Nobody, however, has ever tested if what I wrote was actually true. Therefore, you should try to cite only work that has been peer-reviewed. Another problem with Internet pages is that they tend to disappear. You may find an article today and reference it, but your reader may not be able to find it tomorrow. Academic publications, on the other hand, are properly catalogued and stored in libraries. I suggest you try to reference works that have been published in the following formats (in the order of preference):

  1. Articles in peer-reviewed journals.
  2. Articles in peer-reviewed conferences.
  3. Books. Note that books often contain reviews of other people's works. It is also better to cite a particular part of a book (e.g. a chapter).
  4. Other types of publications (e.g. tech reports, theses, etc).

If you really need to reference an Internet page, then use these APA instructions on how to do it properly.

Finally, we recommend using Harvard style to format the references. In brief, this means using authors' surnames and year of publication, when citing the source in the text, like this (Newell, 1990), and sorting the list of references alphabetically by surnames of first authors:

Newell, A. (1990). Unified theories of cognition. Harvard University Press.

You can see more examples here. Refer to the APA manual for more details.

Top

Plagiarism

Plagiarism is when you are using parts of somebody else's work without acknowledging it. Even using a picture or a diagram from a web-page without acknowledging the author is plagiarism (not to mention quoting text without acknowledging the author). Digital computers and Internet made it very easy to copy and paste information from one document to another, and, unfortunately, some students think they can copy and paste parts of other people's works. These students, however, may not realise that it is also very easy with the help of computers to identify what they have copied and from where. Universities, including Middlesex, are now using special electronic services that can automatically detect plagiarism. So, be aware of that. Plagiarism is a serious form of academic misconduct, and it will not be tolerated. So, don't even think of it.

We value original work. Even if your report contains a lot of quotes that are properly acknowledged, it is your text that will be assessed. So, don't overuse quotes, it is your thoughts that matter. Also, never use copyrighted materials without asking permission from the author (i.e. pictures, diagrams from other sources). If you cannot obtain the permission, then reproduce the item yourself (e.g. reproduce the figure using your software) and say that it is "based on..." and give the reference.

Top

Finally

Your project should demonstrate that you have indeed learnt something in the university and can apply your knowledge and skills to solve new problems. It should also demonstrate that you can manage the project within limited time and deliver results when needed (hence the deadlines). So, don't expect an easy ride, you will have to spend a lot of time and efforts. However, if you pick the topic wisely such that it is interesting to you, then you may find your research exciting and rewarding. The sense of achievement can be even more important to you than the grade itself.

Good luck!