The State of Search 2.0
How useful is search for you? I’m more than dissatisfied with it.
I’m not the only one who thinks we are in definite need of Search 2.0. Just ask the founders of Wink, Delicious, Eurekster, or Rollyo. Or Google and Yahoo! for that matter.
There generally seems to be two different approaches to Search 2.0. The first is the approach of Wink / Delicious, which is largely community driven.
Wink is people powered search. Users rank and tag Google search results, thus using collective wisdom to improve the results for queries. Delicious does something different but with the same goal of better search results - a search query for a word like blogs returns results for the most tagged item by the community.
On the other hand, we have Rollyo a concept based on individuals creating their own search engines. With Rollyo, for example, you can create a vertical search engine on a niche subject or one consisting solely of the sites you visit on a regular basis.
Eurekster is somewhat of a mix of these two ideas. Like Rollyo, Eurekster lets you create your own search engine, calling your creation a “swicki”. In this approach though, you train the swicki to help point it to more relevant sites or search queries. It then uses that information to weight search results - results do not include solely hits from the specified sites. Eurekster also learns from users and has a buzz cloud below its search box, showing the most popular queries by users. You can check out my swicki or see it in action on the right sidebar of Jeff Clavier’s site.
Yahoo! and Google jumped in on Search 2.0 by doing things like social bookmarking and search history. They seem to be betting on personalization over the community approach.
A pure technology solution is not going to solve the search problem. As I wrote in the past, there is no perfect search algorithm. The analysis above proves that. Now, it’s a matter of seeing which type of human intervention - at the individual or community level - brings us closer to Search 2.0.
February 7, 2006 at 10:21 am
Many of the community driven search engines as you call them rely on tags, and bookmarks as opposed to individual voting. At Seekum we allow our community to vote each site up or down depending on how the user feels on relevancy. We believe that as time passes the masses will bring in the best results.
February 10, 2006 at 10:30 am
“A pure technology solution is not going to solve the search problem. As I wrote in the past, there is no perfect search algorithm.”
Very true. One of the biggest problems that I see with search is that there’s too much to sift through already, and the amount of information is still growing extremely fast
. While Google uses things like PageRank to help relevancy, that doesn’t really solve the problem. When I search for something, what may be relevant to lots of other users may not be relevant to me. Thus, what’s relevant to me may be on page 50 of a Google Search.
I’ve come to really appreciate community-based systems like del.icio.us. With a system like del.icio.us, I know I’m not getting ‘everything’ available, but I know I CAN see what lots of other people thought was ‘the best’.