Microsoft's "mind reading" patent

Microsoft applied for a patent to monitor user's brain waves to understand how good (or bad) their interfaces are. Filed on August 9, 2007, the patent application, #20070185697, describes a method of classifying EEG data in a way that separates the wheat from the chaff. From the patent application:

When studying how humans interact with computing devices, it is desirable to be able to determine the effectiveness of a computer-user interface, i.e., a user interface. A traditional way of determining the effectiveness of a user interface is to present a computer user, i.e., a user, with a task, observe the user as he or she operates the user interface to complete the task, and ask the user questions before, during, and/or after the task is performed. The observed behavior and answers to the questions are characterized and quantified. The quantified results are analyzed to determine the effectiveness of the user interface.

Cognitive neuroscience techniques can be used to provide a more direct way to determine the effectiveness of user interfaces. A typical cognitive neuroscience technique involves attaching electrical sensors, ie., sensors, to various points on a user's scalp. The sensors are then connected to an electroencephalograph (EEG). One use of an EEG is to sense electrical changes within the brain that correspond to certain brain states. It is possible to determine the effectiveness of a user interface by analyzing a user's brain states before, during, and/or after a user performs a task using the user interface.

Link (via New Scientist)

• [NineInchNachos]

  how can we be sure that this is not a device that tricks the brain into thinking it likes Microsoft products?

• [Sam]

  Interesting. After Microsoft has proven they don’t have an interface design bone in their body, they are going the scientific route and trying to fine tune their products with trial and error.

  I give them props for trying. It would be interesting to see comparisons made between various operating systems.

  I wonder why they can’t just ask users: “How hard was it to perform this task?”

• [kgreene]

  I had the chance to visit the lab of the Microsoft researcher who’s working on this project. His name is Desney Tan, and he was one of Technology Review’s TR35 honorees. Here’s are a couple of articles I’ve written about his research:

  http://www.technologyreview.com/TR35/Profile.aspx?Cand=T&TRID=629

  http://www.technologyreview.com/Infotech/18962/

• [Maggie Leber]

  Hmmph. More obvious

  [Gunn]

  I’m confused. How can they get a patent for something that so obviously doesn’t work?