Identifying user requirements for novel interaction capture S. Tucker, S. Whittaker and R. Laban Information Studies, University of Shefield Shefield, UK We describe methods used to identify user requirements for the design of new technology in the context of the AMI multi-modal meeting assistant. This application is intended to support access to information exchanged in meetings. One of the main problems in deriving requirements for new technology is that potential users are often unaware of the possibilities of new technology, the new tasks it will support or the new methods it will offer for carrying our existing tasks. These users are also known to be poor at reasoning about hypothetical situations, so that asking the directly about what new technology they need is likely to be ineffective. Instead, our approach has been to focus on observations of users current meeting behaviors; both recording meeting conversations and observing behaviors such as note-taking. By combining these observations with post-hoc participant interviews we are able to identify significant problems that people have in remembering and recording information from meetings. We then generate technical designs that are intended to address these problems, and present them to potential system users to gain feedback about how well these tools address their significant problems. Such feedback is necessary because there are often multiple possible technical solutions to the user problem, and we need to determine which of these is the best. We collect this user feedback by having users carry out tasks with working prototypes before having them comment on other prototype variants embodied as pseudo-prototypes. Paper presented at Measuring Behavior 2005 , 5th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands. © 2005 Noldus Information Technology bv