Checking slashdot, they appear to have filters in their comments for things like 'insightful' and 'imformative'. Thinking more along the lines of being able to rate things like the veracity and tact of the response, with automatic filtering and prevalence based on a few metrics.