Microsoft muses mind meld machinations (and "F" see me)

Tuesday's IT Blogwatch: in which Redmond has figured out how to read the minds of its usability guinea pigs. Not to mention an updated list of funny and/or clever exam answers...

Justin Mullins blogs for the venerable New Scientist mag:

Not content with running your computer, Microsoft now wants to read your mind too. The company says that it is hard to properly evaluate the way people interact with computers since questioning them at the time is distracting and asking questions later may not produce reliable answers.


Instead, Microsoft wants to read the data straight from the user's brain as he or she works away. They plan to do this using electroencephalograms (EEGs) to record electrical signals within the brain. The trouble is that EEG data is filled with artefacts caused, for example, by blinking or involuntary actions, and this is hard to tease apart from the cognitive data that Microsoft would like to study.

So the company has come up with a method for filtering EEG data in such a way that it separates useful cognitive information from the not-so-useful non-cognitive stuff. The company hopes that the data will better enable to them to design user interfaces that people find easy to use. Whether users will want Microsoft reading their brain waves is another matter altogether. [more]

Jacqui Cheng wonder if it could have, "Prevented Clippy":

Anyone who has had to perform in-depth usability tests on others knows that people just plain aren't good at describing what they are doing. In fact, it's a commonly-accepted mantra that, if asking someone to describe what he is doing on a particular user interface, it's wise to also record it or have an observer take notes. There are times when the user may say that he is doing something that is the complete opposite of what he is actually doing without even realizing it ... while someone has yet to invent a mind-reading device (which would no doubt be used to enhance people's love lives more than the usability of UIs), Microsoft wants to get closer to it.


Microsoft ... notes that there have been some efforts to filter out extra noise in EEG readings. This filtering is not always effective, though, and can be expensive to perform. Because of this, Microsoft hopes to bypass the whole conundrum of filtering and instead focus on categorizing brain states that would then be applied when performing usability tests. This would involve taking sample data in a controlled environment, analyzing it for typical patterns ... and then categorizing it into different states of "interest".


Ultimately, these brain states would then be applied to data taken during usability tests to analyze how people think when performing various computer tasks. Certain functions could cause much more cognitive activity than others (and in different ways), which would tell researchers how to tweak it to make it more user-friendly. By itself, the EEG data might provide Microsoft a new angle on usability, but maybe not an entirely complete one. That's why this data could be used in conjunction with other usability techniques, such as eye-tracking and screen capture programs, to create a more well-rounded assessment of a user's interaction with a given UI. [more]

David Pescovitz boings:

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. [more]

But Brian William Jones doesn't think so:

I've spent a fair amount of time looking at this problem ... and I can tell you that this is pretty much up in the night kinda speculation. Technical issues of obtaining clean EEG signals in a convenient manner aside, the origins of techniques like this to classify comes from the epilepsy literature, where folks attempt to classify interictal seizure spikes, but also from the sleep and awareness literature ... Granted, there are rather dramatic global state changes that occur with different states of consciousness, but this application focuses principally on the awake EEG. Specifically they quote the P300 signal which the FBI and CIA are using (trying to use) for lie detection, but the problem here is that the P300 is only an evoked potential that simply tells you whether or not someone recognizes an input (audio, visual, tactile etc...). So all questions or inputs into the system have to be crafted to understand that one is looking for an evoked potential with no necessary context in place to explain that evoked potential.


At its very basic fundamentals this is simply a rehash of signal detection theory doing simple Fourier analysis to "classify" brain waves. But the thing here is that there is no science behind using these signals to interpret what one is thinking even with the invocation of Bayesian networks. There are a number of other more promising methods for classifying data that have been in the literature and commonly used by a number of other disciplines that I am surprised have not made it into the EEG literature yet. [more]


The sad thing is that there is actually quite robust science related to human machine interfaces using EEG data, but I suspect those are already patented fairly heavily and this is some sort of effort to patent what essentially amounts to nonsense. [more]

And DrYak agrees:

EEG is completely useless because it lacks the fine resolution (it only measures a global effect) that other techniques like functional magnetic resonance imaging (fMRI) provide. (And fMRI is far from perfect too : it measures the increased blood flow of brain regions that are working and are needing oxygen. It's much more precise than EEG but its much slower. At best you have a temporal resolution around several seconds)

The gold standard would be inserting electrodes directly into the brain but that's completely out of question for the purpose that Microsoft want (it can be used in some complex neurosurgery to help predict the potential function loss in case of tumor removal)


The only problem I see with this patent, is Microsoft coming after some obscure small research group that try to develop a tool to assist clinician in diagnosis and Microsoft attacking them on ground of infringing a technology that automatically reads and interprets EEGs (for example to try to push contracts for machines with windows licenses for potential buyers of the technology). i.e.: once again restricting development. [more]

If hostguy2004 is to be believed, we shouldn't worry about that:

There are numerous patents relating to EEG technology. I work with Quantitative EEG technology as part of my day job. There is definitely prior art for "de-artifacting" EEG data. [more]

Mary Jane Irwin makes the obvious joke:

Guess they've seen the jagged EKGs of users trying to install Windows Vista. [more]

Buffer overflow:

Around the Net Around Computerworld Previously in IT Blogwatch

And finally... Clever and/or funny exam answers [some new, others old-but-good]

Richi Jennings is an independent analyst/adviser/consultant, specializing in blogging, email, and spam. A 20 year, cross-functional IT veteran, he is also an analyst at Ferris Research. You too can pretend to be Richi's friend on Facebook, or just use boring old email:

Copyright © 2007 IDG Communications, Inc.

Shop Tech Products at Amazon