If we are to learn the effects of the environment and our day-to-day actions, and choices on our physiology, we must develop systems that will label biophysiological senor data with contextual information. In this thesis I first present an architecture and implementation of FEEL: a system for the acquisition, processing and visualization of biophysiological signals and contextual information. The system comprises a mobile client application (FMC) and a backend server, The mobile client collects contextual information: phone call details, email reading details, calendar entries, and user location at a fixed interval that is transmitted to the backend server. The backend server stores the contextual information and biophysiological signal data that is uploaded by the user, processes the information and provides a novel interface for viewing the combined data. Next, I present the results of a 10-day user study in which users wore Electrodermal Activity (EDA) wrist sensors that measured their autonomic arousal levels. These users were requested to upload the sensor data and annotate it at the end of the day at first, and then after two days. One group of users had access to both the signal and the full contextual information collected by the mobile phone and the other group could only access the bio physiological signal. At the end of the study the users were asked to fill in a System Usability Scale (SUS) questionnaire, a user experience survey and a Toronto-Alexithymia (TAS-20) questionnaire. My results show that the FEEL system enables the users to annotate bio-physiological signals at a greater effectiveness than the current state of the art. Finally, I showed that there is a correlation between a person's ability to determine their own arousal level and their score on the Toronto-alexithymia test: the less alexythymic they were, the better their correlation between the EDA and their self-reported arousal.
展开▼
