Authors: Siska Fitrianie, Leon J.M. Rothkrantz
Addresses: Man-Machine Interaction Group, Faculty of Electronics, Mathematics and Computer Science, Delft University of Technology, Mekelweg 4, 2628 CD Delft, The Netherlands. ' Man-Machine Interaction Group, Faculty of Electronics, Mathematics and Computer Science, Delft University of Technology, Mekelweg 4, 2628 CD Delft, The Netherlands; Netherlands Defence Academy, Faculty of Technical Sciences, Den Helder, The Netherlands
Abstract: During crisis events, human operators in a crisis centre will be overloaded with a large amount of information. The stress of dealing with crisis situations can have a significant impact on the certainty of the information. The need for a system that is able to handle information calls automatically may then become apparent. Our research is about a dialogue system that can play as a crisis hotline dispatcher. The dialogue system offers a natural user interaction through its ability to start a user-friendly dialogue taking care of the content, context and user|s emotion. It retrieves information about crisis situations from users while controlling the communication flow. The system is able to recognise the emotion loading of the user|s linguistic content. The recogniser uses a database that contains selected keywords on a 2D |arousal| and |valence| scale. Its output includes an indication of the urgency of the information regarding the crisis.
Keywords: online crisis management; human-computer dialogue; natural language processing; text-based emotion recognition; emergency management; crisis hotline dispatcher; information retrieval; communication flow; information systems; emergency response; human-computer interaction; HCI.
International Journal of Emergency Management, 2008 Vol.5 No.1/2, pp.123 - 144
Published online: 15 Aug 2008 *Full-text access for editors Access for subscribers Purchase this article Comment on this article