Buzz Report, which monitors about 11 million blogs and 100,000 public message boards, already analyzes posts to determine writers' genders and ages. The new release will also measure such factors as geographic location, ethnicity, and income level. Those new demographics will be rolled out over the course of several months, said Dave Howlett, vice president of product management.
Currently, Umbria assesses gender and age by employing a natural language analyzer, which examines a user's posts over time, and constructs information about the person based upon the words and syntax they use. Howlett declined to provide information on how often the algorithm is accurate.
Umbria will launch an automated system of filtering to deal with blog spam--that is, "fake" blogs that are not necessarily representative of the buzz going on in the blogosphere.
Previously, Umbria had been using a largely manual whitelisting and blacklisting approach to blog spam, but found the method too cumbersome to monitor the growing number of blogs. "For companies like ours who are mining the blogosphere, it could easily skew results if you don't get control of it," he said.
Finally, in order to make the various reports delivered to their clients by the Buzz Report more user-friendly, the company pared down the number of reports--from up to 1,900 different kinds of report types to a much more manageable 45. Also along the lines of usability, the Buzz Report includes a weekly and monthly top-line executive summary of the data that is most relevant to the client's brand, and to the brands of their competitors.
Umbria is expected to announce the upgrade later this week.