1
$\begingroup$

A new tag ant-colony has been introduced. I see 2 problems here:

  • The name should have been ant-colony-optimization
  • The tag swarm-intelligence should already cover ant colony optmization techniques, so it does not make much sense for a new tag.

What are your thoughts?

$\endgroup$
1
  • 1
    $\begingroup$ I do favor "ant-colony-optimization" for being more explicit. $\endgroup$ – DukeZhou May 29 '19 at 18:45
1
$\begingroup$

I introduced this tag because ACO is a well developed sub-field of swarm intelligence, so it deserves (IHMO) its own tag, like e.g. reinforcement learning deserves its own tag (compared to machine learning) on a website dedicated to AI.

I used because it is shorter and there's no ambiguity in the field of AI. Furthermore, a lot of people do not refer to these algorithms as "ACO", but e.g. as "ant colony system" or just "ant colony algorithms". I would argue that is a more general tag and expression.

Furthermore, several questions on ACO have already been asked on the website.

$\endgroup$
3
  • $\begingroup$ It's overkill since unlike RL there aren't much questions about swarm intelligence..And also following this trend will result in bee, fish, whale, termite, etc etc optimization. In fact elseiver or some other journal I cannot recall banned the use of 'some random animal' optimisation technique. $\endgroup$ – user9947 May 28 '19 at 5:37
  • $\begingroup$ @DuttaA "there aren't much questions about swarm intelligence". What's the threshold for creating a tag? For me 1-2 questions are enough. The other tags will be created once 1-2 questions on the topic are asked. $\endgroup$ – nbro May 28 '19 at 9:25
  • 1
    $\begingroup$ I personally like this tag! Hopefully we'll get more questions related to ant intelligence, since ants are a very successful species, and extremely good at building tunnels and pathfinding, despite having tiny brains.) $\endgroup$ – DukeZhou May 29 '19 at 17:53
0
$\begingroup$

Wondering if we should maybe just have a general tag for "ant-intelligence" that could cover all aspects, including swarm intelligence, engineering (tunnel building) and path finding..

$\endgroup$
13
  • $\begingroup$ @nbro thoughts? $\endgroup$ – DukeZhou May 29 '19 at 17:57
  • $\begingroup$ I do not think it will be correct, since ants are not the only animal showing collective intelligence. I just thought creating a new tag for just 1-2 questions is not proper $\endgroup$ – user9947 May 29 '19 at 17:59
  • $\begingroup$ @DukeZhou "Ant intelligence" doest not seem to be a common expression (AFAIK), so new users maybe would be lost. I think that ant-colony is a good tag too. ACO is a sub-field of swarm intelligence. As I had said, since there already 5-6 question on the topic, I decided to create it (to be more specific). $\endgroup$ – nbro May 29 '19 at 18:10
  • 1
    $\begingroup$ @DuttaA but ants do seem to have unique qualities, and have been studied in relation to several types of problem solving. $\endgroup$ – DukeZhou May 29 '19 at 18:21
  • $\begingroup$ @nbro Looking at the questions, at least two seem related to pathfinding though... Does "ant-colony" cover this aspect? $\endgroup$ – DukeZhou May 29 '19 at 18:22
  • $\begingroup$ @DukeZhou the thing is that so does bees and termites... Actually some reputed international journal banned the use of 'random animal' optimization. Otherwise researchers were discovering some kind of collective intelligence in some animal and naming it 'that animal ' intelligence. $\endgroup$ – user9947 May 29 '19 at 18:26
  • 1
    $\begingroup$ @DuttaA I hear you. But I can't see that having a specific tag for ant, or even bees or termites, causes harm. (Because ants are popular, it might even give us a little boost!) $\endgroup$ – DukeZhou May 29 '19 at 18:28
  • 1
    $\begingroup$ @DukeZhou just raised a concern..It's subjective though, cause I thought it was not really necessary, not that it's harming anyone. $\endgroup$ – user9947 May 29 '19 at 18:32
  • 1
    $\begingroup$ @DuttaA it's always useful to have these discussions! $\endgroup$ – DukeZhou May 29 '19 at 18:34
  • 1
    $\begingroup$ @DukeZhou Yes, it is true that a lot of these questions are related to path finding. However, given there is a substantial amount of literature on ant-based algorithms, again, I thought about creating this tag. Maybe we will not create specific tags for other sub-fields of swarm intelligence. It will depend on the number of questions we will receive. For now, we don't have too many questions, but maybe in the future. $\endgroup$ – nbro May 29 '19 at 18:39
  • $\begingroup$ I am wondering though if maybe DuttaA's suggestion of "ant-colony-optimization" is helpful in that it makes the tag more explicit. (Less concerned about the length than clarity.) Also wondering if we should expand the definition to include all forms of ant optimization, referencing path-finding, so it works as an umbrella tag in the near-term. $\endgroup$ – DukeZhou May 29 '19 at 18:45
  • 1
    $\begingroup$ @DukeZhou The current description is: "For questions related to ACO". I think that includes all applications, e.g. path finding, but if you want you can add examples of applications to the description. This will not hurt. $\endgroup$ – nbro May 29 '19 at 19:19
  • $\begingroup$ aaah. Just looked at the wiki, and it does state this explicitly. However, this may illustrate the problem--I have a general interest but hadn't initially considered it under this label, in that I was thinking about pathfinding outside of the colony structure... $\endgroup$ – DukeZhou May 29 '19 at 19:22

You must log in to answer this question.