Communicating conceptual integrity in distributed systems through intelligent assistance

Increased globalisation and de-regulation in the commercial environment has given rise to requirements for a new generation of software tools capable of supporting complex, dispersed inter-organisational projects and decision making. A major difficulty in designing such systems relates to the fact that the ultimate users of such tools are commonly from varying backgrounds, have different conceptualisations, frequently use different terms and have varying perceptions of the events, procedures and objects represented within the system. The design of online user assistance for such systems is frequently passive, responding only to requests from the user for help. The author highlights the need for interactive online assistance capable of functioning both in `reactive´ and `pro-active´ modes, with the system instigating interaction when necessary. In particular, it is, therefore, concerned with discussing the appropriate timing and content of the `triggers´ necessary to proactive functioning. A set of design principles for an adaptive assistant which are derived from earlier work on intelligent tutoring system design are described (B. Cox et al., 1988, 1989)