Font Size: a A A

Properties of behavioral decision functions for autonomous negotiation agents

Posted on:2007-09-09Degree:M.A.ScType:Thesis
University:University of Waterloo (Canada)Candidate:Qi, JiaFull Text:PDF
GTID:2448390005975476Subject:Systems Science
Abstract/Summary:
Electronic negotiation is increasingly becoming an important competitor to its traditional face-to-face counterpart. As a result, researchers have formally studied electronic negotiation, and presented various models for its implementation. While the properties of these models have been compared through simulation, an analytical study of the models is absent in the literature. In this thesis, we take up the analytical study of the properties of a specific type of negotiation model, namely the time-dependent tactic (TDT) behavioral model, in which the offers corresponding to the negotiating issues are set to vary with time, according to pre-determined behavioral functions. We study the influence of various parameters of the TDT on negotiation outcome. We then conduct a simulation to evaluate the extent to which our analytical results extend to a range of negotiation scenarios. Our theoretical results not only essentially explain the previous simulation observations about the influence of individual parameters of the TDT, but also provide new insights about multi-parameter interactions that have not been studied in the literature.
Keywords/Search Tags:Negotiation, TDT, Behavioral
Related items