Distinguished Lecture by Professor Shirley Gregor

14 May 2018 at 13:15-15:00 in Lecture hall 2, Ekonomikum, Kyrkogårdsgatan 10, Uppsala

A Theory of Communication Strategies for Intelligent Systems

Shirley Gregor, Australian National University, Australia; Alexander Mädche, Karlsruhe Institute of Technology, Germany;Stefan Morana, Karlsruhe Institute of Technology, Germany

Abstract

Continuing advances in technology mean an increase in the sophistication of interactions between humans and machines. There is a continuing need for effective intelligent system interfaces (ISI) that enable a knowledge dialogue between humans and today’s smart machines that offer both opportunities and threats to organizations and society. The importance of ISI has been recognized since the early days of decision support systems and they have ongoing significance in applications including recommendation agents, advanced analytics, conversational agents, and social robots.

Despite the importance of ISI, the related research remains scattered and the potential for integrating and systemizing knowledge across different forms of ISI has not been sufficiently realized. New forms of ISI are being developed without the benefit of lessons from the past or from experiences with other members of the same class of systems. Thus, the goal of the current work is to initiate development of a Theory of Communication Strategies for Intelligent Systems (CommSIS) that shows how similar communication strategies can be pursued to achieve similar goals across intelligent system interfaces as a class of systems. The goals to be achieved include effectiveness, influence, manipulation, compliance with norms, agreement, transparency and user affective responses including trust. CommSIS is based on a framework from Habermas’ Theory of Communicative Action complemented by a synthesis of empirical studies.

The paper has theoretical significance in that it is an initial step in design theorizing that integrates knowledge, past and present, for an important class of systems. The establishment of this theory allows for the comparison of ISI strategies of different types and their potential for achieving desired goals. The design principles developed in the theory provide a fertile base for the design of ISI in future research and practice.

Biography

Shirley Gregor, Professor of Information Systems, Australian National University, Australia

For general enquires, please contact Deqa Farah-Asbury.