Identify Emergent Trends Based on the Blogosphere

Information about upcoming trends is a valuable knowledge for both, companies and individuals. Detecting trends for a certain topic is of special interest. According to the latest information over 200 million blogs exist in the World Wide Web. Hence, every day millions of posts are published. These blogs contain an enormous think tank of open-source intelligence. Considering the continuously growing nature of the World Wide Web a primary factor of success is the ability to include the latest data and focus on the complete data set of blogs. The structured as well as unstructured data of blogs are available offline via a single database for further analyses. This paper describes and evaluates an algorithm to detect trends based on the data published in blog posts.