Increasingly many systems are being conceptualised, designed and implemented as marketplaces in which autonomous software entities (agents) trade services. These services can be commodities in e-commerce applications or data and knowledge services in information economies. In such systems, dynamic pricing through some form of negotiation or auction protocol is becoming the norm for many goods and customers. Thus, negotiation capabilities for software agents are a central concern. Specifically, agents need to be able to prepare bids for and evaluate offers on behalf of the parties they represent with the aim of obtaining the maximum benefit for their users. They do this according to some negotiation strategies. However, in many cases, determining which strategy to employ is a complex decision making task because of the inherent uncertainty and dynamics of the situation. To this end, this thesis is concerned with developing bidding strategies for a range of auction contexts. In this thesis, we focus on a number of agent mediated e-commerce settings. In particular, we design novel strategies for the continuous double auctions, for the international trading agent competition that involves multiple interrelated auctions, and for multiple overlapping English auctions. All these strategies have been empirically benchmarked against the main other models that have been proposed in the literature and, in all cases, our strategies have been shown to be superior in a wide range of circumstances. Moreover all our models exploit soft computing methods, in particular fuzzy logic and neuro-fuzzy techniques. Such methods are used to cope with the significant degrees of uncertainty that exist in on-line auctions and we show they are a practical solution method for this class of applications. In developing such strategies we believe this work represents an important step towards realising the full potential of bidding agents in e-commerce scenarios.
展开▼
