An Exploration of the Utility of GSR in Locating Events from Personal Lifelogs for Reflection

Digital personal lifelogs (PLs) enable many artifacts from a person’s life to be automatically stored in a digital archive. These data sets can contain a wealth of potentially valuable information describing events from an individual’s life. A key challenge for lifelog technologies is how to develop scenarios and applications which enable people to interact with these vast heterogeneous data sources in a meaningful way. One of the areas where individuals can gain from interacting with lifelog records of their life is in the process of self reflection. To date little attention has been given to applications which automatically extract content from lifelogs to support self reflection using lifelog content. One of the significant issues with reflection from lifelogs is discerning material which may be of interest in reflection from among the huge amount of available data. One way of determining the user’s engagement with their situation is measuring their biometric response associated with their arousal level. Specifically it is known that an individual’s galvanic skin response (GSR) can vary with their level of arousal. We hypothesize that situations of marked GSR variation are likely to be more significant for self reflection than other moments. We present an initial investigation, using 3 subjects’ lifelogs, of the utility of lifelog items with marked GSR for self reflection. Our results indicate that GSR records may serve as a good enabling technology for applications supporting self reflection and awareness.