No Searching for the Swicki: It's Here

 
 
By Clint Boulton  |  Posted 2007-12-04
 
 
 

The swicki is out of beta.

The swicki, formally launched Dec. 4 by search software specialist Eurekster, is a custom search portal and social search widget. The software, free for consumers, also boasts new automation features to make building a swicki simpler.

Users may customize a swicki search portal on any topic, choosing content from blogs, Web sites, images, video and RSS feeds, and share the widget to a community of users.

However, instead of returning millions of results like a generic search engine, swickis "learn" from the search behavior of the community, theoretically making it easier for the next user to find what they're looking for.

Here's how it works: Swickis scan the data indexed by multiple search feeds plus all additional sources specified by the swicki builder to provide relevant results.

The swickis take into account keyword, clicks, votes and behavior recorded from every search, so the search results ranking constantly changes. Swickis also automatically update themselves based on the user's online behavior.

Read more here about Ask.com preparing for collective intelligence.

The new features automate the building process. Now, when users set up their swickis, they can enter their site's URL to trigger auto detection of relevant content for the swicki buzzcloud-a tag cloud of popular search terms-and list of related Web sites to be prioritized in search results.

Swickis can be displayed as rectangular search widgets on Web sites that include a search box and keyword terms added during customization, or as URLs on Web sites.

Eurekster CEO and co-founder Stephen Marder said the company will next add enhanced voting tools and commenting capabilities, and other features requested by users.

Eurekster swickis compete with tools from social search startups Lijit and ChaCha Search, as well as Google's Custom Search Engine and Yahoo Search Builder.

However, while Google's and Yahoo's tools use only their own indexes, swickis pull information from a combination of sources and indexes.

Moreover, swicki search results employ collective intelligence, taking into account every previous keyword search, click, vote and user behavior to help users find what they're searching for.

"The community-driven results and participation by way of voting on the results [something Google is just now experimenting with] makes the swicki more than a custom search experience; it becomes a content discovery engine," Marder told eWEEK.

Top-line search engines Google, Yahoo and Ask.com are adding collective intelligence to their arsenals, but it's taking a lot of reworking of their algorithms-and therefore time-to make this practice effective. Eurekster has done what they want to do, albeit on a much smaller scale.

While in beta, more than 100,000 swickis were created by more than 25,000 people. Those interested may check out how to build a swicki here.

Check out eWEEK.com's Search Center for the latest news, views and analysis on enterprise search technology.

Rocket Fuel