NAAS: Negotiation Automation Architecture with Buyer's Behavior Pattern Prediction Component

Abstract: In this era of "Services" everywhere, with the explosive growth of E-Commerce and B2B transactions, there is a pressing need for the development of intelligent negotiation systems which consists of feasible architecture, a reliable framework and flexible multi agent based protocols developed in specialized negotiation languages with complete semantics and support for message passing between the buyers and sellers. This is possible using web services on the internet. The key issue is negotiation and its automation. In this paper we review the classical negotiation methods and some of the existing architectures and frameworks. We are proposing here a new combinatory framework and architecture, NAAS. The key feature in this framework is a component for prediction or probabilistic behavior pattern recognition of a buyer, along with the other classical approaches of negotiation frameworks and architectures. Negotiation is practically very complex activity to automate without human intervention so in the future we also intend to develop a new protocol which will facilitate automation of all the types of negotiation strategies like bargaining, bidding, auctions, under our NAAS framework.