# How to name an algorithm in a paper

In a paper, I should refer to an existing algorithm in different parts of the paper. There is no name for this algorithm. For the firs time I wrote "the Algorithm proposed by (Xia, et al. 2008) .....", but I don't think it is a good way to use this method any time I want to refer to it. I may name it Algirhtm A, or Algorithm 1 or Algorithm Xia, Algorithm base ...

What do you suggest for referring to this algorithm for several times?

• Calling it "Algirhtm" sure makes it uniquely identifiable ;-) Commented Dec 1, 2016 at 15:57
• FWIW, my MSc thesis references multiple times "the Baeza-Yates–Perleberg approximate string matching algorithm" so I would say using the surnames is an OK option. Commented Dec 1, 2016 at 17:18
• @AndreaLazzarotto, you just did that to increase the word count and page length though, right? :) Commented Dec 1, 2016 at 17:31
• @mikeazo, at most that would've given me 2 more lines, not even pages. :P :D Commented Dec 1, 2016 at 20:10
• @CaptainEmacs I could have several algorithms in a paper. Commented Dec 2, 2016 at 13:09

When you first refer to it say something like "the Algorithm proposed by (Xia, et al. 2008) which we refer to as XAB" where you replace XAB with the first letters of the authors' last names in the order that they appear in the bibliography. So if the authors are Xia, Alpha, Beta, you would use XAB. If the authors are Xia, Yen, Ping, you would use XYP. I wouldn't use more than about the first 3 or 4 author's last names, however.

This way if you are referring to multiple algorithms, you can give a clear reference to each. If it turns out that multiple algorithms you are referencing would have the same name given this scheme, you can also throw in the last two digits of the year. So you would have XAB08 and XAB09.

• Thank you. It's also possible that the same authors proposed another algorithm in another year, and your last suggestion works for this case too. Commented Dec 2, 2016 at 11:14

If you're going to be referencing a single critical algorithm in your paper, I recommend that you simply restate it with a citation, and then reference it by number as you would any other algorithm:

...the algorithm devised by Bob and Joe [ref]:

a^2 + b^2 = c^2        (1)

Now, if we assume a^2 is a cat, then (1) can be extended as follows...

• An algorithm is not a formula. Who is going to mark an algorithm with "(1)"? Commented Dec 1, 2016 at 17:17
• @AndreaLazzarotto - Good point, I was thinking of something more along the lines of signal processing filters and techniques. This clearly won't work in many scenarios. Commented Dec 1, 2016 at 17:43
• @AndreaLazzarotto right, however if one copy the algorithm pseudo-code in the paper, he can refer to it with Algorithm 1, similar to Table 1. Commented Dec 1, 2016 at 18:08
• @Ahmad, ah yes if you put it into a float then absolutely. Commented Dec 1, 2016 at 20:11