Expand AU Menu

Kogod News

 
  • RSS
  • Print

Kogod’s Melander Obtains Patent for Web 3.0 Invention

Art: Nicole Melander invented a way for computers to analyze the meaning of text on the Internet, not just the text itself.

Nicole Melander invented a way for computers to analyze the meaning of text on the Internet, not just the text itself. (Graphic courtesy of Nicole Melander)

Sometime in the future, the very subject of this story might be responsible for the way you found it.

On March 24 Nicole Melander, executive in residence at the Kogod School of Business, received a patent for a process that could one day play a large part in how people use the Internet.

It’s all highly technical and very complicated, but the gist of it is this:  Melander and her research partner, Ophir Frieder, invented a way for computers to analyze the meaning of text on the Internet, not just the text itself.

Melander’s method combines Web coding known as XML, argumentation theory, text analysis, and text encoding standards.

“Let’s think about a court case, something that has pros and cons,” she said. “There’s a standard approach to doing that, certain ways of writing about it. In XML you could notate things as ‘for’ and ‘against.’ You could basically analyze a piece of text that was related to some sort of an argument or a review and see what the supporting evidence was.”

Melander related another example.

“If five different people wrote movie reviews, you can use this method to bring them all together and give the consumer that information in one sort of snapshot,” she said. “We use this to take a piece of text and figure out what the arguments were, and we’re able to present them in a consolidated way.”

It’s part of Web 3.0, or the so-called “Semantic Web” of the future. Now, Internet search engines rely on keywords, which Melander describes as “rudimentary,” to conduct searches. When you Google “turkey,” the computer doesn’t know whether you want information on a country or a bird.

Melander’s method, and others like it, will be able to begin discerning the meaning of text.

“If somebody writes for example, a health document, they tend to write it in a certain structure,” she said. “Once you understand that structure, you can get pretty smart about what you present to somebody who’s going to consume that information.”

Melander’s patent is owned by Oracle, for whom she worked when she finished this research 10 years ago, and it’s unclear just how they plan to use it. The patent process was laborious, to say the least, as she and Frieder had to continually prove to patent officials that their technology was unique.

“Because we combined four different areas, some which individually existed, the patent process was incredibly arcane,” she said. “Receiving the patent was pretty exciting because the process took so long, but also it was a validation of the research.”

Melander won’t be getting rich off her invention, but knowing that her work may play a role in the way people use the Internet in the future certainly is enriching.