I'd like a way to find questions that might be of interest to me, that are not neccesarily marked with one of my "Interesting" tags.

Maybe something that mines viewing history like iTunes "Genius".


The idea is to correlate similar user viewing behaviours in order to provide "suggested reading". I'm presuming that the data dumps don't allow us to see who viewed what questions, due to the anonymity requirement.

We can already see which questions get the most views, fewest answers, and so on using existing features. If I want to answer questions I can browse under Unanswered (as several people have usefully suggested) for questions tagged with subjects I'm interested in. Most of the existing views are sorted lists, with the sort based on some metric (views, votes, etc.).

Perhaps there might be questions (even ones with answers) that would be interesting to me, but they don't have that many views or votes in total - they are in the long tail. But if a few users with similar viewing history to me look at those, then they would be valid items for the "suggested reading" list, especially if not in my preferred or ignored tag sets.

For the blissfully unaware, Apple's iTunes has an opt-in feature ("Genius") that uploads media play information to Apple central. Some algorithm then collates that against similar data from other users, and comes up with suggested items to purchase from the store.


What do you think?

  • 8
    Can you describe what iTunes' Genius does exactly?
    – Pekka
    Mar 24, 2010 at 11:45
  • Unanswered tag should be it.
    – random
    Mar 24, 2010 at 11:57
  • I just earned my tumbleweed badge. I agree with random, go for the unanswered tag first.
    – user50049
    Mar 24, 2010 at 12:43

1 Answer 1


Do the same sort of relationships of what movies and music you like exist with what programming questions/answers you'll like? I doubt it. I think that random is right and you should just browse unanswered. That's what introduced me to jquery which I'm now learning. If you're insistent on having a tool, try coding something that suits your self, maybe as a GreaseMonkey script.

  • 1
    I like movies with lasers, questions with lasers and music with lasers. I think there might be a pattern here.
    – Welbog
    Mar 24, 2010 at 12:09
  • @NlfE Right, but do such patterns extend to programming questions beyond I like python, or I like answering beginner questions. Can it be predicted what tags you like besides your interested tags and your most common tags?
    – C. Ross
    Mar 24, 2010 at 12:20
  • @C. Ross - to answer your question - I don't know, without access to the data its a 'hunch'. In general though it must be true that people who look at similar stuff to me look at similar stuff to me, whatever the subject matter. I was thinking of suggestions for specific items of interest rather than sets (i.e. tags). Mar 29, 2010 at 23:25

You must log in to answer this question.

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