DIADEM

     

July 2010: Paper accepted and presented at IHCI'10:



Andi Winterboer, PhD
Intelligent Systems Lab
University of Amsterdam
Science Park 107, 1098 XG Amsterdam
The Netherlands
Web: www.awinterboer.com

A paper titled 'Mobile Agents in Crisis Situations – Adapting Information to User’s Affective State' by the Diadem members of the University of Amsterdam had been accepted as a full paper at the 4th IADIS Conference on Interfaces and Human Computer Interaction (IHCI’10). At the end of July, we travelled to Freiburg, Germany to present the paper.

The paper is concerned with the effect of emotions on humans’ cognitive information processing and decision-making, which could be very relevant for developing human-computer interfaces in the DIADEM domain, i.e. in crisis-response settings.

The context of this work is an autonomous decision-making system that uses information from people through their mobile phones to assess the nature, location, and seriousness of environmental incidents (e.g., gas spills), predicting the incident’s future development based on, for instance, meteorological data, and alerting people by sending information about new developments or rescue instructions, if required. If such an autonomous system bases decisions on data from ‘human sensors’ it should be able to interpret information from its users as reliably as possible. The knowledge that a user is currently in a certain affective state (for example, sad) could mean that she is likely to provide information in a specific way (substantively) allowing assessing the reliability of information provided by this person more effectively. Similarly, knowledge about the user being in a certain emotional state allows the system to send instructions in a style congruent to a user’s emotional state to increase compliance with the system’s instructions and thus human safety.

The paper reports the results of an 2x2 between subject survey experiment (N=91) with affective state (happy vs. sad) and information presentation style (heuristic vs. substantive) as dimensions. The results confirm that participants in a sad affective state are more likely to comply with mobile agents’ advice when information is tailored to a substantive processing style. They base decisions on substantive information and provide longer descriptions. In contrast, people in a happy affective state prefer heuristic information.

Based on the results of this study, we recommend that mobile agents, especially those interacting with members of the general public in potentially emotionally laden contexts, should adapt the way they present information to the affective state of their users.