Learning and reaction algorithms for certain decision functions used in electronic negotiation

点击翻译按钮获取中文摘要

Despite its downturn in 2000, electronic commerce (e-commerce) remains an essential and growing part of the global economy, with studies indicating an annual increase in sales of over 20% during the last few years.  One change that is brought about by e-commerce is in the area of negotiation. Researchers have designed a number of systems, known as Electronic Negotiation (EN) systems, which use computer networks to facilitate the exchange of information/offers between the negotiating parties. Owing to the rapid adoption of online markets, EN is becoming an increasingly important competitor to traditional face-to-face negotiation.  A steady stream of research literature has attested to the improvements afforded to human negotiators on account of learning an opponent's preferences.  However, the development of learning methodologies for the EN context is somewhat in its infancy.  This research will propose methods for an EN system to learn an opponent's negotiation tactics and to use that knowledge to react in an optimal manner.  We will apply our methodologies to specific business situations, and also start to develop a framework that can be used to design robust EN systems.  The development of EN systems will be important to the e-commerce industry, and this research will provide for the advanced training of personnel in an emerging area.