Spacilus is a new interesting application which acts as a proxy and tries to categorize Google search results in order to eliminate unwanted items (spam, scraper sites, etc).

Spacilus uses Bayesian statistics to categorize search results. Put simply, Bayesian classification looks at the joint probabilities of a search results words and phrases to belong to a certain category. For example, a search result that contains a phrase like “price comparison” will most likely belong to a category “Commercial” or “Spam” rather than to a category “Informative”.