Group items tagged

Get songs, videos, tour dates and leave a personal note to the Prince of Darkness on the fan wal

Since the start of the year, a team of researchers at Carnegie Mellon University — supported by grants from the Defense Advanced Research Projects Agency and Google, and tapping into a research supercomputing cluster provided by Yahoo — has been fine-tuning a computer system that is trying to master semantics by learning more like a human. Its beating hardware heart is a sleek, silver-gray computer — calculating 24 hours a day, seven days a week — that resides in a basement computer center at the university, in Pittsburgh. The computer was primed by the researchers with some basic knowledge in various categories and set loose on the Web with a mission to teach itself.

The Never-Ending Language Learning system, or NELL, has made an impressive showing so far. NELL scans hundreds of millions of Web pages for text patterns that it uses to learn facts, 390,000 to date, with an estimated accuracy of 87 percent. These facts are grouped into semantic categories — cities, companies, sports teams, actors, universities, plants and 274 others. The category facts are things like “San Francisco is a city” and “sunflower is a plant.”

Edmunds.com, which supplies consumers with information on new and used cars, provides an example. IT workers at the firm originally began using Splunk to better understand patterns of visits to the website—for example, to determine whether a sudden spike in traffic was a cyberattack or the result of an article that went viral—and to help track system logs. Today Edmunds uses Splunk for much more. "We now provide data dashboards to business groups in the company to show them up-to-date information that helps them make decisions," says John Martin, senior director of application operations at Edmunds. "The information that business groups need is in our IT logs, but it wasn't being looked for." One such dashboard allows the editorial and marketing staff to track in real time which auto brands and models are attracting the most views. This allows the business to experiment with ways to be more responsive to what users are doing hour by hour. For example, ads could be shuffled to match what people are most interested in at any moment.