Real-time search
The web is constantly being updated with new information. Search engines like Google, Yahoo, etc. index this info into an archive and then search that archive to bring you results. But what about the content that doesn’t get indexed in time? Real-time search is addressing this issue.
Search start-ups such as Collecta, Scoopler, OneRiot and Yauba (to name a few) aggregate content being shared on the internet as it happens by continually indexing updates on sites such as Twitter, Digg, Facebook and more. In fact, Microsoft just announced that they’ve integrated Twitter and Facebook posts into Bing search results as they take their first step into real-time search.
As interesting as real-time search promises to be, I’m curious how it will handle how to accurately rank (does it appear in spot 1 or 100) a piece of content that has just been added to the internet. How do we know that the results real-time search serves up are relevant to what we are searching for, other than having a related keyword in the search query? Traditional methods of ranking search results such as link authority (whether or not trusted links are pointing to this content) won’t really work with real-time search, as content just added to the web is not going to have inbound links from trusted sources.
One of Bing’s approaches to ranking real-time search results (at least for Twitter) is to assign higher value to longer tweets, those that contain added information and links, and tweets from users with many followers. While that approach sounds good in theory, it seems like something that could be manipulated fairly easily.
It’s pretty exciting to see how real-time search continues to unfold. What effects do you think it will have on how we search for things online?
