What should be a master's thesis in computer science?

St. Petersburg Academic University continues to enroll in the magistracy . Hurry up to apply, the first interviews have already passed.

In this post, we will discuss what should be the master's thesis of computer science.

Master theses: bad and good

Successful training in the magistracy inevitably ends with the defense of the master's thesis.
Let's discuss how a good master’s thesis is different from a bad one.
')
Our department protects two types of master's theses: theoretical (theoretical computer science and bioinformatics) and applied (software development).

With theoretical master's, everything is arranged simply: a good master's thesis is a scientific article adopted at an international conference. For example, this post describes one of these theoretical master's theses (link to an article in the comments) that our graduate defended (after which he left for a PhD at the University of New York). The recognition of the results by the international scientific community is a very good and convenient indicator of the quality of the master's thesis.

Much more complicated is the case with applied master's. This is due to the fact that in the evaluation of applied master's degree is much more subjective. The master's thesis is a research qualification work devoted to solving an actual problem, having theoretical or practical significance for modern science and technology. From here follow the three main evaluation criteria, as well as the three most important questions that need to be able to respond to the defense of a master’s thesis: novelty, relevance and practical importance.

Let's deal with each of them.

1. Novelty
   Everything is simple with the novelty: the result described in the master's diploma should contain something new. To show that the result is really new, you need to analyze the existing solutions of the task and show what the difference between the proposed solution and the existing ones is. If the master's work simply repeats something that has already been implemented before (for example, “OOP library for working with the network in C ++”), then this work does not contain any novelty.
2. Relevance
   The problem, which is solved in the master's thesis, should be significant enough for the present moment and interesting in itself. Ideally, of course, it would be a big plus if someone has already tried to tackle this task, and the undergraduate has managed to improve the result. In practice, more often it happens that someone has already engaged in a similar task. In this case, you need to explain why the problem being solved is also relevant. If the statement of the problem is taken from the ceiling, such as “minimizing the number of commas in the C code”, then its relevance is very doubtful. At the same time, the task itself may be non-trivial and very difficult, but only its solution does not interest anyone.
3. Practical meaning
   An applied master's thesis should have some kind of applied application, i.e. The result should be not only new and important, but also necessary. For example, the work “ABC language compiler in XYZ language” may be new and relevant, but completely inapplicable, for example, if the given compiler works unrealistically for a long time.

If the master's thesis does not meet these requirements, then such a master should not be allowed to defend. Unfortunately, in practice, a significant part of master's degrees, which are defended in Russia, do not satisfy these simple requirements. It is quite possible to defend a master's degree on the topic “Implementing an Information System for a Library Using a DBMS” and even get 4 or 5 for it. If this is possible in your university, then you should think about the appropriateness of training in it.

Who is to blame and what to do?

Who is to blame for bad master's happenings? In most cases, the entire responsibility lies with the university, with the exception of those rare cases when the student is a dolt, and "flunked" the writing of the master's degree. Here, of course, the question immediately arises, how did it happen that this dolt reached the 6th year?

In other cases, the task of the university to help the student find a manager for writing the master's, control the process of writing it and notice in time if “something went wrong.” In our master's program, we are very attentive to the master's work of our students. Firstly, we help to find managers, under whose leadership you can write a good master's degree. Secondly, during the sixth year, we repeatedly arrange extended pre-defenses and discuss with the students the problems they have, teach them how to properly represent the results of their work and answer tricky questions. This way you get work like this ( part 2 ) or this .

We asked some of our graduates of this year to write articles on the results of their master's degrees.
Soon we will publish in our hub.