16

Duplicated questions are a thing we want to avoid on these sites. However, it will happen anyway, this is why we get the "exact duplicate" close reason.

There are some questions which are coming back frequently, and they are closed as duplicate of an initial one, usually. For questions which are coming back the most, it could be good to take over the original question to improve it (because often it was badly asked, compared to a duplicate, which is finally a bit silly, but it has the most answers). Eventually, it could be nice to even make some CW, so that they become references.

Besides the fact of "Is it a good idea to take over a question someone else asked, for the greater good?", there is the question of "how to find these?". I know that so far, the link to duplicate is only included in the question, which is potentially editable, so it's not a solid reference.

Besides this constraint, is there a way to find which are the questions which are the most referred to in a duplicate context?

I know it is not possible with the current search tools, but maybe it could be with a data dump? As a further idea, would this have an interest in being formalized (like for example including the main duplicate link to the close description), and eventually have a "top duplicates" accessible in moderation tools?

|improve this question
  • 1
    I for one would like this, to improve the quality of the most referred to question and close all related ones on sight – Ivo Flipse Feb 9 '10 at 9:54
  • 2
    This should definitely be available in the moderation/10k tools. – Ether Feb 9 '10 at 18:33
  • I'd like to see more data in the data dump in order to produce queries like this. Please vote for this related feature request: meta.stackexchange.com/questions/37524/… – Greg Hewgill Feb 9 '10 at 21:09
5

The improved close-as-dupe page shows the similar posts and count of posts linked to as the duplicate to the similar post.

new-improved!

|improve this answer
8

These only count those dupes where the "possible duplicate" link(s) where inserted automatically; this feature hasn't always been in the site. I just used a very simple matching algorithm, too, so take the numbers with a pinch of salt. Only shows the top twenty (for Server Fault only those with more than one dupe). Based on the February data dump.

Meta:

Stack Overflow:

Super User:

Server Fault:

|improve this answer
  • Awesome, it's funny to see what questions are apperently asked over and over again – Ivo Flipse Feb 9 '10 at 10:32
  • 3
    Time to create a new list of "fave" questions on SU. – random Feb 9 '10 at 10:40
  • Thank you very much, that's great! – Gnoupi Feb 9 '10 at 10:49
  • Another super-common dupe that I swear I see every day: "what's the hardest code you've ever written?" – Ether Feb 9 '10 at 18:34
  • 2 and a half years on, and I would love to see the top dupes for SO... Is there a data explorer query for this? – Oded Aug 9 '12 at 20:53
4

This is a very good idea!

I can see a big potential for this. For instance on Super User, there are like a gazillion questions about "syncing 2 folders" and dupes come in almost every day. I still need to look for the original though. So having a list of all these kind op "mother-dupes" would be awesome.

|improve this answer

You must log in to answer this question.

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