5

Thanks to large amount of meta-data available on the SE network, data mining seems to be a popular activity, so much so that it was a sponsored activity on Kaggle awhile back. No surprise then that research (even if not peer-reviewed) is coming out using this data:

Paper and Executive Summary

Want a Good Answer? Ask a Good Question First!

Data Mining Reveals the Secret to Getting Good Answers

Abstract:

In this paper, we study the problem of inferring the quality of questions and answers through a case study of a software CQA (Stack Overflow). Our key finding is that the quality of an answer is strongly positively correlated with that of its question. Armed with this observation, we propose a family of algorithms to jointly predict the quality of questions and answers, for both quantifying numerical quality scores and differentiating the high-quality questions/answers from those of low quality. We conduct extensive experimental evaluations to demonstrate the effectiveness and efficiency of our methods.

Discussion*:

What do you think about the validity of the conclusions and methodology employed in the paper? Is there anything useful that we can use to make the site better, or does it simply state the obvious?

* For the record, I have no involvement with this paper nor the authors. I thought that the meta-community at-large would be interested.

Improve this question
4
  • 5
    Wow. "Quality" based only on post scores...
    – Mat
    Commented Dec 9, 2013 at 6:52
  • 3
    Higer-scored question does not mean it's better. To get some quality comparison, you should compare it to the average and standard deviation of it's tag, at least. There are SE sites for math and for statistics, right? So there you can get details, but naked vote count is a poor measurement. Also, they tend to ignore the fact that vote count depends also on the SEO position of a question, both in internal and external search. "Ask a popular question to get a popular answer" could be a title of this just as well.
    – Mołot
    Commented Dec 9, 2013 at 8:14
  • "For instance, if we simply predict everything as low quality, we would get a high prediction accuracy."
    – Brad Larson Mod
    Commented Dec 9, 2013 at 15:30
  • While they control for only the answers that come in during the first 24 hours, it's unclear to me if they also only account for votes that come in during that first period. The data dumps generally don't make that easy to extract, and I know that question and answer voting has a very long tail. If they're not accounting for that, the question and answers scores from older questions will skew these results.
    – Brad Larson Mod
    Commented Dec 9, 2013 at 15:32

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .