2 ms·
The results come from the real-time community actions and from the Google index you use everyday. Basically, Webzzle is to 'explore' what Google is to 'search'.
by webzzle 17y ago
The results come from the real-time community actions and from the Google index you use everyday. Basically, Webzzle is to 'explore' what Google is to 'search'. We all use Google to search but Google is using keywords. It's good when it's the first step of your search process but not convenient when you want to know more from a Web page. What keywords will you select ? At Webzzle, we analyse the Web page from which you want to know more (chen you click 'explore' and send one query to Webzzle (to get Real-Time knowledge from the community) and many queries to Google in order to help you find the best results in 1 click instead of multiples keywords queries by yourself in the Google search interface.
- baddox 17y agoI for one am still confused. I went to the wikipedia page for "disco" and hit Explore. It took me to a webzzle page with a link to a disco band and a list of top disco songs, then a google "enhanced results" list which was basically the standard google results for "disco" but with only results related to music. For example, the Mac burning software "Disco" is #2 on google, but didn't appear on the webzzle results. So I understand that it's working by not showing non-music-related results, but I still don't grasp the purpose. Please give an example of a usage of this program, for example: while at pageA I clicked Explore because I wanted to see X and these results helped me out.
- webzzle 17y agoYou start to get it. We basically use Wikipedia content to build the semantic Web and try to dramatically improve the search results quality. We build the meaning based Web. Keywords don't handle the meaning. The purpose of the explore feature is to save time for the user in the search results analysis process and get higher quality results : go straight to better results. As the community grows, the Webzzle results improve. It's the knowledge network effect.