The biggest problem with Athena is that it needs the right training data to be effective. If we can find a way to match a wiki with its best training data that would be ideal. It would matching wikis based on similar communities and content styles.

This is obviously no simple task, and would require lots of planning and is no where near a priority.