An adaptive negotiation strategy for real-time bilateral negotiations

Abstract

Each year the Automated Negotiating Agent Competition (ANAC) introduces an increasingly complex negotiation setting to stimulate the development of negotiation strategies. This year, the competition featured a real-time bilateral negotiation setting with private reservation values and time-based discounts. This work introduces the strategy of one of the top three finalists: The Negotiator Reloaded (TNR). TNR is the first ANAC agent created using the BOA framework, a framework that allows separately developing and optimizing the components of a negotiation strategy. The agent uses a complex strategy that takes the opponent’s behavior and the domain characteristics into account. This work presents the implementation, optimization, and evaluation of the strategy.