Smart Apps Become Smarter

Monday, June 6, 2016 @ 04:06 PM gHale


A smart application can now recognize and reason about a human’s activities in order to be able to coach in a purposeful way.

Esteban Guerrero at Umeå University in Sweden has new computer-based methods based on activity-centric and argument-based theories.

RELATED STORIES
Hacking into a Hacker’s Character
Random Numbers could Hike Security
Security Software? Think Again
Ensuring ‘Trustworthy’ Systems

In daily life, humans evaluate his or her activities based on more or less explicit information. A skier uses information about arm- and leg movements, distance, environment, etc. based on the goals the person has set up. But there are also factors that affect that are less explicit, such as the motives behind improving health, social inclusion, etc.

Guerrero’s research aimed at developing theories and methods, which also include complex factors in the computations of capacity and performance. When more complex factors end up included, the methods need to handle uncertainty and changing conditions.

By using and developing theories based on human activity and reasoning such as activity theory and argumentation theory, different interpretations of a situation can end up generated and evaluated, and adjusted when new information comes in, so the human recognizes it and can participate in it.

Guerrero started with assessment methods used by therapists, among other an instrument for measuring balance and strength in older adults for preventing falls, and developed new generic methods that a computer system can use. These methods build on activity-theoretical models of human activity and on new-developed argumentation-theoretical frameworks.

The methods ended up implemented in mobile apps tested among other older adults to evaluate capacity and performance in exercises that aim at measuring different aspects of strength and balance. This was done in collaboration with physiotherapists at the department of Community Medicine and Rehabilitation at Umeå University.

“The methods could be used in for instance in ‘smart homes,’ for example diagnosis and treatment apps that the person can use at home, or an app measuring and evaluating balance and strength for preventing falls in older adults,” Guerrero said.

Guerrero has performed his graduate studies within the User, Interaction and Knowledge Modelling research group at the Department of Computing Science at Umeå University.