Exactly one year since the beta launch of one of the Web’s hottest realtime search engines, OneRiot is dropping the “beta” tag and reemerging as version 1.0. Along with a site redesign marking its first birthday, the search site is also getting a powerful new tool implemented directly into its algorithms.
The Realtime Web Trending Topics Engine is OneRiot’s latest attempt to break stories and surface trends before any of the competition. Founded on OneRiot PulseRank technology and realtime index of the social Web, the Trending Topics Engine uses a complex system of social site analysis (of Twitter, Facebook, MySpace, Digg, and a OneRiot panel), natural language processing, weighting, filtering, and, most importantly, clustering.
OneRiot will cluster certain topics closer together depending on their weights, and each cluster gets scores based on their “general hotness” (general popularity) and “emerging hotness” (dependent on rate of acceleration across the Web).
“Everyone is familiar with Trending Topics on services like Twitter and Yahoo,” explains Tobias Peggs, President of OneRiot. “One of the primary modes of activity on the realtime web is ‘content discovery’ and clicking on trending topics drives a lot of that activity. Specifically on realtime search, this is one of the primary uses – an amazing 80% of the search queries we see at OneRiot are users looking to track today’s trending topics, to find out what’s going on right now.”
There are many reasons why realtime search is so important. Besides content discovery, the single most important aspect of all this is the effect it has on advertising. OneRiot says click rates are astounding when tied to trending topics, which makes a ton of sense: in the end, it comes down to relevance to the user.
