Modelling variations in goal-directed dialogue
Jean Carletta · 1990
Introduction This research investigates h. uman dialogue variations by having simulated agents converse about a simple map navigation task using a computational theory of human dialogue strategies. The theory to he tested is that agents have a variety of strategies which they use in goal-directed dialogue for deciding what information to include, how to present it, how [o construct references, and how to handle interactions. Agents choose strategies for each of these aspects of the dialogue depending on the complexity of the current task, responses from the partner, a. nd the task's importance; in general, agents try to minimize the collaborative effort spent on he task but still complete it satisfactorily. The software allows sinrelated agents to be constructed using any combination of strategies, showing how the strategies interact and allowing the decisions made in human dialogues about the same task to be modelled. Currently, the system works on a subset. of the strategies found i