Back in October 2016, there was an announcement that Stack Exchange would be collaborating with the University of Melbourne on duplicate question detection.

It looks like the portion of work where we provide input ended in January.

What is the result of the work we did? Will Stack Exchange be able to utilize this work on any (or all) of the sites?


Thanks to Julia Silge @juliasilge tweet we know that there are publications related to this from the author.

Some relevant work is here and the paper here.

| improve this answer | |

Thanks for your interest! The thing that Stack Exchange was most interested in was how to recognise when two posts are falsely labeled as duplicates.

According to Stack Exchange this is one of the biggest frustrations of new users. They take the time to write a question, and within minutes it is flagged as a duplicate. This is fine when it really is one, of course, but when it's not, it's disappointing, and can make people leave the forum, never to come back.

A paper on this topic is currently under review. If it gets accepted, then Stack Exchange can decide to use the findings if they want.

| improve this answer | |
  • 1
    Any update on the paper's review? – Monica Cellio Oct 25 '17 at 20:35
  • Any chance you could post the paper to e.g. arXiv while it's under review? – Jeffrey Bosboom Oct 26 '17 at 0:25
  • 3
    Hi Monica and Jeffrey, the paper was rejected, but we worked on it some more and resubmitted it somewhere else. I'm not sure if I'm allowed to but it on arXiv, but I can definitely post a link to it here if it gets accepted. – Monozygotic Oct 27 '17 at 0:04
  • @Monozygotic Was the paper ever accepted? I have no idea how long the process takes. – Thunderforge Jan 12 '18 at 19:13
  • 2
    @Thunderforge, I submitted the paper to a journal, which can take quite a long time to respond. I'm still waiting for a reply. It would be great if they could reply within the next month, because I'm finishing my PhD, but unfortunately all we can do is wait. – Monozygotic Jan 14 '18 at 22:25
  • @Monozygotic Is the data available and the preprint public? Were we can find more about this? – llrs Aug 27 '18 at 15:35
  • 2
    "They take the time to write a question, and within minutes it is flagged as a duplicate" because in almost all cases it is a duplicate. – Raedwald Mar 21 '19 at 13:15

You must log in to answer this question.

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