Approximating user values to preserve privacy – a proposal

Users have different sets of personal values, such as benevolence, self-direction, and tradition. Among other factors, these personal values influence users’ emotions, preferences, motivations, and ways of performing tasks and hence, information needs. We sketch a method where, during software development, multiple value-dependent interface variants with different functions are created. When used the first time, personal values of the individual user are identified, and the software presents itself in the variant that best matches these values. In this paper we focus on identifying values when using software the first time. Currently used methods to identify values are work intensive and/or solicit personal user information. A method intended for routine use when a user starts using the interface, should require little effort and not intrude privacy. Instead of probing for user values directly, we propose to approximate users’ personal values based on the users’ preferences for work tasks and to neglect other factors influencing preferences. Questionnaires allow efficient data collection, and users have few issues sharing opinions about work. Inasmuch as this indirect querying of user values approximates underlying values, appropriate interfaces can be identified when using the software.