This is the second of a three part series on how the Darwin Awareness Engine™ addresses some of the key problems generated by information overload on the Web. The Awareness Engine is a Web aggregation utility for anyone needing to maintain awareness of emerging events from the Internet and intranet. The technology provides a temporal semantic visualization of events, by means of a Scan Cloud™ that facilitates a rapid understanding of trends, outliers and anomalies.
Issue Two: Popularity versus temporal correlation ranking:
Search engines operate on popularly ranking to deliver information. This method is rooted in the Web 1.0 where information existed in great quantity and had a longer shelf life. The Web 2.0 explosion generates information that is often isolated and relevant for a short period of time. This type of information may only appear in the 100th-plus page of search engine results. In contrast, our temporal correlation-ranking model seeks the latest published information regardless of its popularity. This new model considers that the themes associated with the information are noteworthy if other related information is emerging across a defined timeframe. This suggests an emerging or accelerating movement. Darwin Awareness Engine reveals this emergence through its BuzzTape and ScanCloud.
Example: The BuzzTape on Darwin Awareness Engine’s General News Edition displayed the relative acceleration of the term “Lou Dobbs”. Clicking on “Lou Dobbs” displayed a ScanCloud created from over 200 informal Web 2.0 social events showing terms such as “racist”, ”Latino Americans”, “petition”, “fire”, “CNN” all created within the last 2 hours. We searched Google News for the following five days without any coverage of this social reaction against Lou Dobbs.
While being specific in our search, by adding terms that our technology correlated, we discovered related Web 2.0 documents buried deeply into the search results. It took one week after noticing the social event for the press to address the issue and make the announcement of Lou Dobbs’ resignation over his indiscretions towards the Latino community. Note that, in this example, we never programmed the system to specifically target Lou Dobbs. The emergence was detected by our technology, as a result of his name being a term with great acceleration over just a few hours.
In the next post we will look at organic semantics
Comments