Does http archive store some kind of category (e.g. business - ecommerce, information technology etc) for the URLs being used.
Would be useful to look at the collected data category wise (instead of rank wise) - may offer some useful insights/similarities.