Tailoring activity recognition to provide cues that trigger autobiographical memory of elderly people

Abstract

About a 19% of elderly population is associated with poor performance in assessments of memory; the phenomenon is known as Age-related Memory Impairment (AMI). Lifelogging technologies can contribute to compensate for memories deficits. However, no matter how functional technology is, older people will not use it if they perceive it as intrusive or embarrassing. This paper shows our work to tailor current activity recognition techniques (based on Emerging Patterns) to provide value for AMI people from RFID reading and GPS positioning. Evaluation shows (1) increases in the recall of autobiographical memories, (2) recognition issues, which require the supervision of the e-Memory Diary, and (3) evidences that this approach didn’t suffer from the usual rejection showed to this technology by elderlies.