What is Non-Linear Search?
I was asked about non-linear search and said I’d give it a go.
(The question comes from Simon Collister, who I am sure has a few ideas of his own up his sleeve. But since he wrote a fab post about tagging, which in turn fuelled my own effort on the subject, it’s definitely my turn to go first!)
Non-linear search is one of the bounties of the web 2.0 approach that has been relatively unheralded because of the noise surrounding the ongoing “digg/wikipedia/myspace/youtube is heaven/hell” wrangles.
The expression comes from a post on the BusinessWeek blog. Interviewing Joshua Schachter from del.icio.us, Heather Green notes:
Joshua made an interesting distinction. Instead of finding information a la Google, social search is about finding knowledge. The idea is how do you connect with the information you need in a context that’s knitted together by people and by human expertise, rather than the linear way we do it now, which is to type a search term into a box.
So what do we understand by linear and non-linear search in this context?
Linear search: You already know the answer in many senses. Or you know exactly what you’re looking for and there is a finite answer. What is the capital of Uzbekistan? (”capital city Uzbekistan”). Even quite tricky facts are accessible. Room rates at the Hilton Hotel in Paris? (hilton paris “Rates” -”Paris Hilton”)
A lot of other searches work less well, though. “Good hotel London” brings up a load of sites that want to sell me rooms. What I really wanted, though, was a recommendation of a good hotel in London that meets my criteria - pretty central, cheap, not horrible. How am I going to find that on Google? How do I find “romantic restaurant cheap rome”, “what are the best blogs about supermarket wine” and “inexpensive ways to make your wife feel special”?
So how can I do better?
Non-linear search: The same search terms about the cheap, central-London hotel on del.icio.us produced this:

A site with customer reviews, a Google site that might help me judge their worth, and a specific recommendation. So which was more useful? (rhetorical)
Back to the idea of knowledge. A lot of the questions we really want to ask are the questions that we’d ask another human being, if we knew a person with the right qualifications for the job. Finding the name ‘Tashkent’ is one thing, but what is it like there? Where should I stay? What things should I see? Is there anything or anywhere I should avoid?
I’ve got no idea how to find the answers to those questions using Google. Simply entering the country’s name into del.icio.us provides me with a lot of better ideas:

These sites provide me with some contacts. Some real people. If I were going there, I’d ask the authors of ProgressoR and Registan, or find a message board there. And I’d probably ask Craig Murray (the UK’s ambassador to the country, as it happens), since he’s a fellow brit. I somehow think they might be more use for my questions than the CIA or Jimmy Wales, the top links on a Google search.
The Yahoo Answers site provides another solution to this quandary, though obviously, you’d need to rely on a resident of Uzbekistan (or wherever, or whatever) being a willing and able answerer. I haven’t tried it, but I’m not sure I fancy my chances on Uzbekistan nightlife. Depends what you want to know - they would probably do better on London hotels.
So how do tags help? Tags are a way of finding people who speak the same language as you. “Good cheap hotel London” means something real to those people - the people who use my language know what I mean to a far greater extent than Google ever could. When I search on tags rather than keywords, there is a better chance of finding them.
It all depends what you want from the Internet. A really big library/encyclopedia - the way they used to describe the web in the 90s? Fine - I think we’re probably already there. But I also think we can possibly expect a bit more. Immediate contact, perhaps, with specialists in whatever field you’re researching. Access to knowledge as well as facts.
People, networks, tags, wisdom - these things, for me, are 2.0. Forget the buzz startups and believe the promise.