I have been studying the concept of developing web communities and since i experienced 'stackoverflow like' Q&A communities, i decided to focus on concepts of Q&A communities. On a specific issue, a Q&A community is based on knowledge, for sure.
Knowledge-based/Rule-based Systems in classical AI use predicate inference to create either a yes/no answer and/or a probability associated with said answer(s). The "crowd sourcing" of askbot, more resembles the behavior of Google PageRank in the sense that answers are weighted by reputation, both behaviorally and mathematically. Unlike PageRank however, the Q/A system here is not a "zero sum game". In PageRank, the universe of URLs must equal a constant, and so the weight of site links is constantly recalculated, so no one can "cheat" in the search unlike many other search engines. Askbot does not subtract karma just because others' karma increases -- it is not a curved grade, it is a respect-based grade, or an "honors" based system if you will.
As you mentioned, the basic concept is knowledge. People ask new questions only if they believe that they can get really useful information. This requires having an expert group as a prerequisite in the community. Novice people can increase the traffic and question database.
But the first focus of the community should not be on novices, it should be on experts. Because if there is a sufficient number of experts in the community, then novices will follow them. But the reverse is not true, namely experts won't follow a community web site full of novices.
Comments