I haven't used the service much, mainly focusing on semantic clustering. Like many other services, it takes a little getting used to the interface. As a query clusterer, I really liked it. It truly makes working with semantics easier. Additionally, I've heard that it's used for analyzing or recovering drops. I haven't worked with that feature yet, but I'll keep it in mind. The reviews about the service are all positive.
