GHC12: New Investigators - Mobile Devices

Moderators: Gilda Garreton (Oracle), Rachel Pottinger (University of British Columbia)

 On User Privacy in Personalized Mobile Services

 Michaela Goetz, Twitter, presented her GHC12 Award Winning paper,

Ms. Goetz's paper covered her research on how to best target advertising without compromising user privacy - tricky if you want to find out which advertising is working and to target advertising appropriately.

Her research included a method for doing this without requiring a trusted third party server, which involves doing counts by including noise terms - enough to protect the privacy without statistically impacting the overall counts.

It's nice to allow the users to set their own personal context for what is sensitive or not: for example, going to the hospital to see a relative would be very private, but walking the dog may not be so. While it's nice for the advertisers to learn what is sensitive or not, but even then they could learn more about the user than desired.

Understanding How Children Use Touchscreens

Presented by Quincy Brown, Assistant Professor, Bowie State University.

This is an important topic, as millions of these devices are being sold. Kids love these touch screen products, but they are not designed with little fingers and skill sets.  Children have trouble with things like dragging - the concept of maintaining contact to drag was surprisingly challenging.

Dr. Brown's research covered adults as well, on several different devices. In one of the experiments, they measured success based on target size (area size you could touch in order to get the desired action) and gesture interaction (how they could do things like drawing letters or symbols). Children miss the targets twice as often as adults and found (unsurprisingly) that larger targets were easier for the children to find.

The researches discovered a new phenomenon: holdovers! When the application was "too slow" to respond so the user wasn't sure if they had hit the target or not, they would repeat their action. 96% of the "holdovers" came from children.

Kids gestures were also different - they lifted their fingers more often. For example, to draw a square, children frequently drew 4 independent lines, whereas adults never lifted their finger, just turning their finger to make the shape. This causes problems for the touch device - it doesn't recognize four lines that overlap at the corners as a "square".

This post syndicated from Thoughts on security, beer, theater and biking!