Odysseus

Lifelogging is a new and exciting area of research concerned with recording aspects of a user’s life in digital format. A lifelog can contain any digital items which a user may have accessed or generated, e.g. emails, web pages or photos. In our work documents are captured from interactions with desktop computers and mobile phones, and images from daily life are recorded using a Microsoft Research SenseCam. This is a small wearable digital camera (worn around the neck) that is designed to take photos passively (without user intervention). When worn continuously a SenseCam will typically capture approximately 3,000 images from across a user’s day.

Lifelogs typically contain very huge amounts of information and an interesting question is whether we can identify items which the user is likely to recall or remember afterwards. People often exhibit measureable biometric responses to activities in their life, for example heart rate and skin temperature can vary depending on their situation. These variations often correspond to their emotional response to a situation or the impact it has on them.  Biometric response can be captured and archived digitally using sensor such as heart rate monitors and specialist devices such as a BodyMedia armband. Using other sensors such as monitoring of nearby Bluetooth devices, it is also possible to monitor aspects of social context, such as whether the user is in a shared social environment or is alone. This project will explore the potential interactions between human activities, biometric response, social context and the likelihood of events being recalled afterwards.

The project will build on existing research activities in the CDVP/CLARITY exploring the relationship between lifelogs and biometric response. The work will contribute directly to our ongoing research into developing tools for managing digital lifelogs digital memory archives and automatic analysis of social networks and contexts.