5

When you have a data science-related question, there are at least seven relevant Stack Exchange sites to choose from. So how do you pick the right one?

Yes, for each question you can first ask on Meta (example: What it is the most appropriate Stack Exchange site to ask about ontologies and knowledge bases?), but that's not the best solution.

Can we create a definitive guide (think of a flowchart) that helps to choose the right SE site?

The list of relevant websites:

  1. Stack Overflow. Stats: 15m questions 23m answers 71% answered 7.9m users 9.6m visits/day 7.5k questions/day 9y3m site age.
  2. Data Science. Stats: 6.7k questions 8.2k answers 71% answered 31k users 11k visits/day 13 questions/day 3y5m site age.
  3. Cross Validated. Stats: 112k questions 108k answers 60% answered 138k users 111k visits/day 93 questions/day 7y3m site age.
  4. Artificial Intelligence. Stats: 1k questions 1.6k answers 79% answered 9.1k users 556 visits/day 3.1 questions/day 1y2m site age.
  5. Computer Science. Stats: 22k questions 27k answers 83% answered 62k users 13k visits/day 21 questions/day 5y7m site age.
  6. Theoretical Computer Science. Stats: 9k questions 13k answers 76% answered 31k users 1.9k visits/day 3.8 questions/day 7y2m site age.
  7. Computational Science. Stats: 6.7k questions 9.1k answers 78% answered 18k users 2.9k visits/day 3.8 questions/day 5y11m site age.

I assume that ideally we want to have only one correct website per question. All the questions asked on the incorrect websites should be closed/migrated to the correct ones. Otherwise we'll have confusion, duplicate questions and scattered user base. I think that a definitive guide would help both for asking new questions and for migrating the old ones.

Related questions:

Improve this question
3
  • 1
    Relevant, maybe a duplicate: Which Stack Exchange website for machine learning and computational algorithms?
    – E_net4
    Commented Oct 18, 2017 at 23:09
  • @E_net4 That question is about specific type of questions within Data Science (Machine Learning algorithms), and I'd like to create a guide for all sorts of DS-related questions. But thanks for pointing out, I've added it to the question.
    – Dennis Golomazov
    Commented Oct 18, 2017 at 23:12
  • There isn't always one right site for a question. In particular, those sites indeed have a significant overlap, ans their domain of on-topicness is unlikely to change soon. Why not just choose to ask at Data Science?
    – E_net4
    Commented Oct 18, 2017 at 23:19

1 Answer 1

Reset to default
3

It's a total mess, and DS is an extremely vague term. Your list is quite long but despite its length misses several other Stack Exchange websites (e.g., linguistics, quants, or maths). Here is my personal approximate flowchart:

  1. if the question is 100% coding issues, post on SO
  2. otherwise: if on-topic on CV, post on CV
  3. otherwise: post on DS

At that point Stack Exchange is so fragmented that I think we need to allow "crossover questions" between sites.

When https://datascience.stackexchange.com/ and https://ai.meta.stackexchange.com got created, there were some fruitless debates regarding the difference with CV:

Improve this answer

You must log in to answer this question.

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