Using real-life troubleshooting interactions to inform self-assistance design
Jacki O'Neill, Antonietta Grasso, Stefania Castellani, Peter Tolmie
End useres facing technical problems with machinery, as for example computers and printers, can be assisted by systems that guide them toward an autonomous solutions of the problem. Systems that can be offered to them are wide in their range, but typically follow either in the category of Expert Systems or in the category of searchable databases that can be queried with keyword searches. Both the approaches present advantages and disadvantages in terms of flexibility to address different levels of user expertise and easiness of maintenance, however little studies explicitly address the issue of how best design the balance between guidance and user freedom in such systems. In the work presented here an office equipement call centre has been studied in order to understand which kind of mechanisms are used when a human agent is guiding the user toward the resolution. The overall aim being not at reproducing the agent behaviour exactly in a system, rather at identifying whcih interaction building blocks such a system should have. A critical analysis of an existing knowledge base is also presented together with the ethographic observation in order to further exemplify the user study findings.
Proceedings of INTERACT, Rome, Italy, 12-16 September 2005.