A company called Paterva (which might be South African) is offering Evolution, a search engine that focuses on people and organizations (and associated data such as phone numbers, email addresses etc.), purporting to reveal connections between people that are significant. It seems aimed at security firms, investigators and the like (opening tagline: "Your new enemy is virtual"). There's a web-based search tool that you can try, and a Java-based tool to download that's in beta but that offers to display relationships graphically (screenshot).
I haven't tried the download but I have searched for myself on the browser-based tool. It raked up a ton of old, forgotten stuff, most of which was utterly uninteresting. And as for phone numbers, it came up with one or two out-dated numbers and a bunch connected to other "Foddens" in the world. I can't say I'm too impressed — but, this does give you a glimpse of what it might, with continued development, turn into, and what the download application might do.
Overall there's something of a scary or even creepy feel about this. (As it's searching, the sources it's banging on flash by in a little graphic beside the search box; you can see "Interpol" whip past.) It is true, as Paterva says a number of times, all of the data is freely available now — and I guess that's the scary thing: how much loose but private information is in fact out there. I haven't done a point by point comparison with the results from a good old meta-search engine to see whether Evolution is in fact scraping deeper and wider or whether a lot of this is shine.
The service might be aiming to improve its offering by leveraging data supplied by users. Your search probably didn't help them much, but if you searched for Britney Spears, they would probably start highlighting that particular connection. Over time they might be able to suggest more about the character of the relationship.
An interesting idea, which of course is entirely benign if appropriately anonymized. Social mapping has been around for ages. The unethical part is that the data is not anonymous, compounded by the fact that they might just be plain wrong in many cases.
Judging by the presentation here , the automated nature of the service would likely produce false hits. That presentation is interesting, though. there is a fairly frank expression of a security-focussed hostility in the voice (its titled "I know what you (and your company) did last summer, on a blood-red background), which you can also hear in the website. A different worldview than Slaw's.