4

I have an issue with the attribution requirements for the Stack Exchange Data Dump. For references, data are under CC BY-SA license and the requirements are :

  1. Visually indicate that the content is from the Stack Exchange network
  2. Link back to the original source question or answer
  3. Display the author names for each question and answer you show
  4. Link back to the author’s user page

I'm planning to create a Markov chain generator based on the data dump. My issues are that apart from point 1, I will have an hard time to comply. By its nature, the chain generator will produce text with source spanning most of the data I have.

My natural reaction would be to indicate the content origin but don't follow point 2, 3 and 4 as it would be impractical. But I'm not sure that's a good solution.

So I was hoping someone with more experience than me could tell how one is supposed to do proper attribution when using the data to train a Markov chain generator.

Improve this question
6
  • So basically you're asking if few attributions rules can be omitted out of 4? Right?
    – HackerKarma
    Commented Oct 10, 2015 at 14:25
  • Also see the post about Microsoft's Developer Assistant for Visual Studio SO content attribution. Microsoft is complying to first two rules. But, it would be great to know if ALL the rules must comply or few are fine.
    – HackerKarma
    Commented Oct 10, 2015 at 14:35
  • 2
    @HackerKarma Basically, yes.
    – gkr
    Commented Oct 10, 2015 at 17:04
  • Did you see the Microsoft post linked in my above comments? SE is fine with them so there is some flexibility but could be changed in future.
    – HackerKarma
    Commented Oct 10, 2015 at 17:44
  • 1
    @HackerKarma yes, I've seen the link. I assume nothing bad will happen.
    – gkr
    Commented Oct 10, 2015 at 20:48
  • You might want to ask about this case on Open Source.
    – unor
    Commented Jan 17, 2016 at 4:34

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .