Experimental modeling of human-human multi-threaded dialogues in the presence of a manual-visual task

It’s Alexander Shyrokov, and this is my first post here. Andrew Kun, Peter Heeman and I wrote a paper entitled “Experimental modeling of human-human multi-threaded dialogues in the presence of a manual-visual task” that was accepted to the 8th SIGdial Workshop on Discourse and Dialogue in Belgium (will be held this September). In short (the paper itself is a short paper, but it’s still four pages long) we were looking at how a person who is engaged in driving (the driver, or as we refer to this person in the paper, the officer) and some ongoing task communicates a new thread to the other person (the dispatcher) involved in an ongoing task. The ongoing task was navigation from one place to another and the new thread was forced by presenting a warning message on the screen of the driving simulator.

The following video demonstrates the experiment we ran to collect data. Text in the right bottom corner shows what task is active at any given time. There are three segments in this video. The first two show the driver, and the last shows the dispatcher. In all cases the driver and the dispatcher are involved in the ongoing task and then switch to the interrupting task.