On the flip side of my privacy discussion, I've had reasonable success tracking people down on the net: having done it a couple of times I thought I'd share my successes.
In these few cases, I spent a half-hour or so with google: most of the time, it was pretty easy to do, keeping potential matches and scanning through a lot of chaff, and building a consistent profile of the 'target'.
I wondered what kind of people were in the mod team with me? Starting from their forum handles, I google'd a few. A few had pretty common handles, but one had a pretty unique username. I chucked it into google.
This produced about five or six forum profiles. Most had almost no personal info, but one had a link with a real name, also pretty rare.
Running the real name through google produced some interesting results. Several hits were from a paticular country, which I already knew matched my target's country.
Then I struck gold. Google returned the website for a band, where the keyboard player had the target's name. Even better, there was a picture of the band at a practice showing the target. Some of the info about the band indicated that they were together at college.
Combining one of the band members and the target's name returned a word document containing a college class list including the two names. The country and area matched, so I was pretty sure it was correct.
This gave me the name of the college and even the class ID. Googling again, it returned a handy timetable of the lessons for the class including the lecture theatres, dates, and subjects.
To recap, from an 'anonymous' name and hidden profile on the original site, I was able to find a picture, real name, school and timetable. In theory, I would have been able to stand outside waiting for the target.
(I later sent the photo to my friend in an online chat session - he was extremely surprised as he wasn't even aware the band had a site!)
I was looking for an old school friend, someone who'd been to my primary school. Starting from Friends Reunited, she had an account profile there but hadn't checked there for a while.
On her profile, she mentioned that she'd contributed some research to a paticular wiki article and she now lived in London.
Checking out the wiki page, the history gave me the list of edits. Of the article, most of the text had been added at one paticular point by one user, and all subsequent edits had been spelling mistakes, rewording and grammar correction. I had a pretty good bet that the username who added the main chunk of text was my target.
This username had also been discussing edits on another page, which I found on wiki search. This new page also had edits by the user, and one of them linked to a flickr photo.
The flickr photo gave me a flickr username (different) but checking through the other photos by this user, and reading the comments, confirmed my suspicion this was probably the target.
Google then was able to give me the other places that this flickr username had been used. I concentrated on the forum hits that included photos from the flickr account and found a regularly used forum that contained several of the flickr links. The forum poster username was different again, but shared a common topic linking it to the other usernames.
The new forum was interesting: it was a 'local' forum for a paticular area of London, and was quite happy to give me a list of posts by a paticular user.
At one point, discussion had touched on a paticular local shop, mentioned in other posts by name.
This gave me the exact local area of the poster. Combined with a couple of the flickr photos of 'my new flat' and 'the park across the road', and the googlemap ariel photos, I tracked down the exact building.
I stopped then, because it was getting creepy and felt voyueristic. Total time : 48 minutes (including half-watching Eastenders).
This was a person who did not have a big online presence, used multiple usernames, and stuck to general internet saftey rules (no addresses or real names in profiles).
This method isn't guaranteed, and does depend on building a likely profile of the target, crosschecking with what you know. You also end up wading through lots of chaff, but if your target has a relatively unique name it's a lot easier. You can also cross-ref with the location (town, country) and anything else you know, such as hobbies or interests.