University of Arkansas Computer Scientists Receive Award at Mobisys Conference

Photo Submitted

An app designed by the researchers that visualizes mood and stress of a mobile user. They use this kind of information to adjust the phone's performance to best suit the user's needs.

Tingxin Yan, assistant professor of computer science and computer engineering, and doctoral student Jon Hammer received the Best Poster award at the 2014 Mobisys Conference in June. MobiSys is a leading conference on mobile system research.

Yan and Hammer’s poster, titled “A Virtual Sensing Framework for Mobile Phones,” describes a system they have developed that can sense what a mobile phone user is doing and feeling and optimize the phone’s performance accordingly.

Many mobile applications use hardware to detect user activity. A pedometer can use the phone’s accelerometer or GPS system to detect a user’s movements, for example. But relying on hardware sensors can drain the phone’s battery and use up 3 gigabyte or 4 gigabyte of data.

Yan and Hammer’s system relies only on information that is already stored by the phone, such as which apps are being used, which notifications are being sent and how the user is reacting to notifications. A machine learning algorithm processes this data and prescribes the phone’s response. So if a user is responding to emails but ignoring social media notifications, the phone might use this information to conclude that the user is at work and adjust by making work-related notifications and applications more prominent.

Yan and Hammer have implemented a demo version of their system on an Android phone, and they plan to develop a complete open source system that app developers can use to make their own applications.

“I’m very proud of this team,” said Susan Gauch, head of the department of computer science and computer engineering. “Their work has the potential to change the way people interact with their mobile devices and make these devices even more helpful and intuitive.”