Re-thinking lifelogging : designing human-centric prosthetic memory devices

• Main Author: Kalnikaite, Vaiva
• Published: University of Sheffield 2009
• Subjects: 150.724
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.505524

Building Prosthetic Memory (PM) technology has been an active research area for the past few decades, with the primary aim in supporting Organic Memory (OM) in remembering everyday events and experiences. Through building and evaluating new PM tools, this thesis attempts to explore how and when PM tools are used to help OM in everyday memory tasks. The focus of this thesis is to investigate PM tools as an extension of, or a supplement to, OM and to understand why people choose to use PM as opposed to their OM to help them retrieve information. Further aims of this thesis are to investigate the role of Metamemory and social processes. Finally, the work aims to support Autobiographical memory through building new PM tools. The studies apply mixed experimental and naturalistic methods, and include 3 controlled lab studies and 3 field trials involving a total of 217 participants. Overall, there were 5 new PM devices built and evaluated in long-term and controlled contexts. Results obtained through lab studies suggest that PM and OM function in a synergetic relationship. In particular, use of PM increases when OM is particularly weak and this interaction is mediated by organic Metamemory processes. PM properties also have an influence - people prefer efficient over accurate PM devices. Furthermore, PM cues help in two ways: 1) at encoding to help focus OM; and 2) at retrieval to cue partially remembered information. Longer term studies also reveal that PM is not used to substitute for OM. Instead users prefer to use recordings to access specific parts of a lecture rather than listen to the whole thing. Such tools are extensively used by non-native speakers, although only native speakers' coursework benefits from usage. PM tools that support social summarisation demonstrate that people exploit social feedback and cues provided by other users and that these improve recall. IV Finally, evaluations of new autobiographical memory tools show that people upload mementos based on their importance. There is evidence for preference for mementos that are associated with other people and home. I conclude with a discussion of the design and theory implications of this work.