Does a glossary make sense in a computer science masters' thesis

Question

This question was first asked on writing, a user pointed me at this site.

I am currently writing my master thesis in computer science, more specifically on voice activity detection, a sub-area of automatic speech recognition using neural networks. I asked my supervisor if I needed a glossary and he told me that was up to me. I looked at other computer science works, some have a glossary, some don't. Now I'm unsure on what to do. Here are my thoughts:

On the one hand, a glossary is useful especially in printed documents to quickly find the best explanation for important terms. It provides a proper place for definitions. And to me it provides a feeling of "scientificity".

On the other hand, almost all readers will use my work in pdf form. Thus they can easily search for appearances and definitions on a given phrase. Given my work is only of use to computer scientists I'd consider them aware of these functionalities. Thus a glossary feels a bit anachronistic.

I hope this qualifies as a proper question and isn't too opinion based, I figured rules here might be a little less strict considering the "proper" way of writing something is often influenced by personal preference.

Answer 1

I'd encourage you to include a glossary.

I expect your work to use a mixed vocabulary from different fields of science, e.g. computer science, signal processing, or even linguistics. Your paper might be read by people who are not as familiar with all these fields as you are. And some of the terms might not have a common, generally-accepted definition. Or they might gain a different meaning over time (e.g. "AI" now has a rather different meaning today than it had in its early years).

So, defining and explaining the vocabulary that you use seems quite necessary to me, to make your paper accessible to an audience from slightly different fields of research, and understandable for a longer period in time.

As a (maybe old-fashioned?) reader, I'd prefer to have these definitions in one place, not forcing the reader to search through all occurrences of the word just to find its meaning.

Answer 2

It depends on, first, your advisor, but also on your expected audience. If you think it makes sense and your advisor agrees, then it makes sense. Also, if you define some new terms or use standard terms in a non-standard way, then it might even be more valuable, though, perhaps, not necessary.

If you intend to publish it formally, then an editor of a journal (or the committee of a conference) might ask for a revision. But that is a different issue that you can easily deal with if it arises.

In the absence of constraints, do what you think best. Not a problem.

Answer 3

On the one hand, a glossary is useful especially in printed documents to quickly find the best explanation for important terms. It provides a proper place for definitions.

I'd contest that the proper place is a background chapter that provides a story underpinning what comes next. Not everything need be introduced there, for instance, there's no need to introduce an acronym that'll you'll use in some distant chapter (just introduce it in that chapter), but the core concepts should be, especially since a masters' thesis presumably demands demonstrating knowledge of the fundamentals - maybe ask your advisor about this, it might be a good way to ensure you aren't dropping marks.

On the other hand, almost all readers will use my work in pdf form. Thus they can easily search for appearances and definitions on a given phrase. Given my work is only of use to computer scientists I'd consider them aware of these functionalities. Thus a glossary feels a bit anachronistic.

Searching isn't always as easy as it should me! To make it easier, use italics the first time you introduce a term that you define, it'll help readers find what they are looking for.

---

Some form of hybrid approach might be possible, especially with the number of tools that LaTeX provides for this type of thing.