Multi-modal Opponent Behaviour Prognosis in E-Negotiations

摘要

Automated negotiations introduce a challenging research field that aims to enhance the performance and optimise several aspects of the electronic marketplace. This paper is concerned with the design and evaluation of negotiation strategies suitable for intelligent agents acting in Business-to-Consumer e-commerce environments. In order to minimize the cases where an agreement is not reached upon the expiration of its deadline, the client agent is enhanced with various function approximation techniques, which aim to predict the behaviour of the provider agent during the last negotiation rounds.