Group items tagged

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.”

“This is all about finding a way to give advertisers the opportunities they used to have in the days of three networks,” said Mr. Manning, when marketers could effectively and efficiently reach huge audiences just by buying commercial time on ABC, CBS and NBC. “People are still watching television, but they’re spread out among hundreds of channels and the Internet,” Mr. Manning said. “The one place where people re-aggregate themselves back into a crowd again is the retail store.”

Gilt Taste, the company’s food platform, is to release a free iPad-only app, also called Gilt Taste, on Wednesday. Its 140 recipes are presented straightforwardly — all text, one recipe step on each page, no videos — except for one game-changing feature. Using the iPad’s built-in camera, which tracks your hand movements, you can turn the pages of the recipe without touching the tablet. Lift your hand in front of the screen, brush it from right to left (as if turning the page of a book), and the screen flips to the next step. Wave your hand from left to right, and it goes back to the last step.