I have this doubt.
I've noticed on sites like Yahoo Answers that there is a recognition of the semantics of the questions and they are categorized automatically. Of course, there are bugs, but it is very effective most of the time.
Which method is used?
I've already thought of ways to do it, but I'd like to hear from you here.
I thought about doing a keyword count on the reported text, and thus, direct you to the category that contains those keywords. It would be a sort of "punctuation", where with each keyword found, adds a point to the category that contains it, in a field like "cat_keywords" in the database.
Another question I have is about computing resources. Would such an algorithm not consume many resources?