1

As I understood correctly, in the field of computer science, there is no "standard notation" in scientific publications for explaining various aspects of a computer system. Even though in each discipline there exist some conventions for denoting various elements in the system (e.g., in networking and communication, usually N refers to a node and T is time, etc.), for them to be used in a scientific paper, usually one needs to define them in the paper anyways.

My question here is that, how can I properly continue using an already available and well-defined set of notations, terminologies, and the system model from an already published article in my new paper such that I don't require to re-define or explain them?

To provide a specific example, in this paper I defined a set of notations as well s a system model. If I want to continue using the same notations and system model in my new paper, how can I properly cite it in my current paper and skip re-explaining them without making potential confusion for reviewers?

A real example (with links) would be much appreciated so that I can discuss them with my co-authors.


Update 1: As stated by @DaveLRenfro and @Anyon in the comments, it looks like a common practice in the field of mathematics and physics to refer to a previously published paper for terminologies and notations. This question, though, is concerning more on a similar practice in the field of computer science.

Update 2: Providing evidence regarding reusing terminology, notation, and system models in any academic paper in the field of computer science would be highly appreciated.

Update 3: There are mixed opinions about re-using terminologies from another paper, but in this case, I believe it's preferable to do so as it's in line with the paper that I want to borrow terminologies and more importantly notations from.

Update 4: As discussed in the comments (now moved in the chat), there are over 3 dozens hits in Google scholar for the phrase "for terms not defined in this paper" but almost all of them are either related to the field of mathematics or physics.

1
  • Comments are not for extended discussion; this conversation has been moved to chat. Please read this FAQ before posting another comment. We can only move comments to chat once. Jun 22, 2021 at 16:21

2 Answers 2

3

I'm not sure why you think that for things that require precise definitions that the standard practice should be any different in mathematics and CS (I've studied, published, and taught both).

It may not do your readers a great service, but it is enough to say that your notations, definitions of terms, etc. are taken from a specific cited source. This saves you from any charge of either plagiarism or copyright infringement.

It is a disservice to your readers only if it isn't likely that they have already seen the earlier paper and need to read it before they can begin yours. It is also, in that case, a disservice to yourself as some won't bother.

However, many people will be able to discern much of what you write from the context provided that it is well written.

But, if you cite the original source of your notations, then a brief paraphrase of the meaning of a term at first use would be both acceptable and helpful.

If you want/need to repeat full definitions from the earlier paper then citing, again, solves the plagiarism issue, but you need to consider copyright. Some things can't be copyrighted at all if an idea can be expressed in essentially only one way. The derivative in math is like that. The idea is too bound with the expression to allow copyright of the expression. (Ideas can't be copyrighted - only expression). But in other cases you might need to both cite and paraphrase if too much of the earlier expression needs to be copied otherwise.

And, when you do copy, make sure that you clearly denote what is copied with such things as quotes or indentation.

But, again, the practice of mathematicians and computer scientists isn't really different for such things.

7
  • Thanks for your great answer. I do agree with all the points you mentioned. However, as discussed in the comments section with others, it looks like there is a subtle difference between CS and math when it comes to re-using notations (at least in some specific disciplines of CS). But regardless, do you know any specific example (in any disciplines of computer science) where a paper referred to another one for terminologies and notations? Jun 21, 2021 at 19:33
  • 1
    Not specifically at the moment, but an author might well do so in a follow up paper on one they wrote earlier. Within a small specialty field, the notations are likely to become common knowledge pretty fast since people will read one another's work.
    – Buffy
    Jun 21, 2021 at 19:37
  • It is definitely a lot more common in CS papers to (perhaps needlessly) restate even standard definitions.
    – Maeher
    Jun 22, 2021 at 9:16
  • @Maeher Sure, you are right about definitions. Things like a "computer network" or "RAM", "CPU", "scheduling", "time complexity", etc. are quite common in CS papers. What about mathematical notations? I haven't seen any paper in CS (mostly in computer networking) that refers to another paper for mathematical notations. Jun 22, 2021 at 13:25
  • 1
    @Aruralreader Sure. For example, notations used and presented in my previous paper where I explained an approach for simulating a specific type of cloud systems. Now in my new paper, I'm looking for ways to extend the previous model/notations so that I can use them to explain my proposed method for optimizing a cloud-native system. I prefer not to explain everything from scratch and like to re-use the previous set of notations, if possible. arxiv.org/pdf/2103.08983.pdf Jun 25, 2021 at 14:02
2
+50

In principle, there is no reason that definitions/notations cannot be cited, just like substantive arguments or observations or other parts of a paper. However, it raises the practical problem that the reader has to have recourse to another paper in order to understand the notation and meaning of terms in your present paper. Irrespective of the particular field, the Golden Rule in all such cases is to be clear and make things easy for your reader, so that should be your guiding principle in this decision.

The dilemma you are experiencing only comes about when the definitions/notation are sufficiently volumous that you want to avoid a long repitition of them. So, first have a think about whether there is a way to present the required definitions/notation in a more parsimonious form. Ask yourself: have I reduced this to its simplest form? In the event that you can significantly reduce this material, you might find that it is not onerous to put it directly in the new paper.

If you decide to rely on definitions in a previous cited paper, I strongly recommend that you at least give some paraphrased heuristic summary of what the definitions mean. This is something that is useful to "jog the memory" of a reader who has read your other paper, and if your heuristic explanation is good enough, it might even serve as an adequate substitute for the formal definitions in some cases (allowing a reader to roughly understand your paper even without going back to the previous one).

There is no silver bullet here, so ultimately you will have to decide on the appropriate trade off between parsimony in the present paper, and reader convenience in having a set of notations/definitions in one place. Good luck.

5
  • Thanks, Ben. If I understood correctly, your answer implies that I still have to "re-explain" the notations, terminologies, and the model in the new paper, at least in a minimalistic form. Do you think I should move the explanation, notation table and system model in the appendix or keep it in the main part? Jul 1, 2021 at 8:59
  • BTW, this is the paper that I want to follow notations from: arxiv.org/pdf/2103.08983.pdf Jul 1, 2021 at 9:00
  • You also mentioned: "there is no reason that definitions/notations cannot be cited, just like substantive arguments or observations or other parts of a paper" --> Do you have any example in the file of computer science that a paper cited another paper's notations? As mentioned in Update 4, it's common in physics or math but I couldn't find any example in CS yet... Jul 1, 2021 at 9:12
  • I don't have any strong views on particulars; just follow the Golden Rule and you'll be fine. Sorry, I don't have any examples of this.
    – Ben
    Jul 1, 2021 at 10:35
  • I awarded the bounty to your answer merely because I disagree with @Buffy on "the practice of mathematicians and computer scientists aren't really different for such things." I think, unlike in math, in CS, it's extremely unusual to refer to another paper for notations, therefore better not to do it...I might be wrong, though, but I couldn't find a single piece of evidence. Jul 5, 2021 at 16:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .