Jan 11, 2012

Samsung prototype: Smartphone knows if you're ticked, tweets angry face symbol

Let's say you're having the day from hell, be it dealing with malfunctioning equipment, the server was hacked, or you've talked to one too many technically-challenged customers. Do you trust your smartphone enough to interact on personal touchy-feely level with you? Would you want your mobile phone to read your emotions and know you are ticked, and then react accordingly such as marking your tweet with a frowny-angry face?

We are posed on the brink of computers recognizing and responding to users' emotions; yes, real life HAL. While that might have good uses, it would "not be cool if the computer started closing applications because it judged you were too tired, too frustrated, or too excited to continue." Now emotion-reading technology is coming to your smartphone; your mobile phone will know when you're mad. An example might be that you read an article that further ticks you off about SOPA, since it just infuriates all geeks, and you go to tweet Boycott SOPA with a link to an Android app that identifies SOPA supporters, but the smartphone system determines you are not in perfect control of your emotions. What if your smartphone said, "Sorry, you may not tweet at this time. Take a chill pill, mellow out, and try again later." Sound too far-fetched in the sci-fi realm?
Next week at the Consumer Communications & Networking Conference (CCNC) 2012, Samsung researchers will show off a prototype smartphone system that can detect a person's emotions. Instead of spying on you via biosensors or the mobile phone's camera, it will monitor other types of input like if you are furiously shaking, how fast you are typing, or how often you press "backspace" and other "special symbol buttons."
While tapping a touchscreen as you wildly text may not exactly nail your emotion, it's enough to guess at if you're "happy, sad, surprised, fearful, angry, or disgusted." Hosub Lee, a researcher for the intelligent computing lab at Samsung Electronics told Technology Review that "there are subtle correlations between these behaviors and one's mental state, which the software's machine-learning algorithms can detect with an accuracy of 67.5 percent."
The prototype, an Android-based Samsung Galaxy S II, works with a Twitter client so each tweet will have a symbol to indicate the tweeter's emotional state. A few more tweaks and this technology could be enabled for texting, Facebook, email, or even comments on social networking sites.
But wait, there's more! This smartphone system may go well beyond certain ringtones you might have setup on your mobile phone for special people. Another "potential" application might include ringtones that alert you to the caller's emotional state. If that mood was "Grrrr, I'm ready to bite you," it would give us entirely new reasons to screen calls and be grateful for voicemail. On the other hand, Lee said your smartphone might try to cheer you up with a "funny cartoon to make the user feel better."
To use this technology, each person will go through training at first to honestly record his or her mood at the time of tweeting. The Samsung system will store your mood data as well other factors that affect mood like "weather or lighting conditions." The variables are then "all fed into a type of probabilistic machine-learning algorithm known as a Bayesian network, which analyzes the data to identify correlations between different emotions and the user's behavior and context."
Previously MIT Affective Computing research has shown "how computers can be more emotionally intelligent, especially responding to a person's frustration in a way that reduces negative feelings." Now I'm not sure about you, but if an "emotionally intelligent" computer or smartphone were to determine I was too frustrated or angry to tweet, and therefore would not permit it until I cooled down, then my temper would only tick hotter. However this smartphone isn't meant to prevent social networking, but instead to share emotions; so on a great day, your tweets might all be marked with happy faces. Webroot has software to stop you if you've been drinking and driving a keyboard, and you are about to post something stupid on your social media account; if you can't pass a test in which you must type the alphabet backwards in 60 seconds, then it won't allow you to post.
The Samsung prototype accuracy may not yet be very high, but Lee said, "Emotion recognition technology will be an entry point for elaborate context-aware systems for future consumer electronics devices or services. If we know the emotion of each user, we can provide more personalized services." Ah yes, the better to potentially datamine and invade privacy or to select personalized ads — marketers should love it. 
There are all kinds of logical reasons not to tweet or post anything of a personal nature anywhere online; it makes it that much easier to know precisely what would work best to social engineer you. A person can only speculate what kind of personal vulnerabilities might be revealed when emotions are harvested, or what this might mean for future smartphone attacks, but malware aimed at stealing your reality might become much more successful. And what if your smartphone rats you out, via someone datamining your emotions? How long might your mobile carrier store that personal info for law enforcement?
At DefCon, Michael "theprez98" Schearer presented "WTF Happened to the Constitution?! The Right to Privacy in the Digital Age," in which he noted, "just because we can share, doesn't mean we have to." It's a slippery slope; the more oversharers voluntarily put out too much information on social networks, the more it "lowers society's expectation of privacy and subjects us to more government intrusion."
Reference :Computerworld 


Post a Comment