How is the tagging system implemented here at S[O|F|U]?

Question

We have a home-grown learning management system (LMS - highly abused term, but I'll be damned if I can find a more succinct way of putting it) that includes a list of classroom classes (filed under various categories) and a list of online educational modules (filed under various categories) - and there have been some (justified) complaints about it being difficult to find things.

I've read a couple of articles on creating a tagging system, and have seen a number of implementations. I was curious as to how the S[O|F|U] implementation was set up as it seems to be much better than most of what I've found.

Please don't hear what I'm not saying: I don't want the code handed to me on a silver platter. I'm just curious as to the general DB/Infrastructure used for the tags here as I like the system much better than what I've found elsewhere.

Answer 1

The db model is the easy part, the tricky part is the find-as-you-type implementations which auto-completes tags, it is tricky to get it working properly cause of various browser annoyances (finding where the cursor is, is non-trivial).

On the DB side a PostsTags and Tags table should be enough to get you going.

From a performance perspective, you have to cache total counts per tag and don't want to end up in deadlock central.

Then you may also want to look at Tag synonyms so for example if someone tags a topic with c# you could automatically retag to c-sharp.

Finally you may want to consider admin tags. (tags that can only be applied by admins)