Modeling User Preferences and Mediating Agents in Electronic Commerce

Abstract

An important ingredient in agentmediated Electronic Commerce is the presence of intelligent agents that assist Electronic Commerce participants (e.g., individual users, other agents, organisations). These mediating agents are in principle autonomous agents that will interact with their environments (e.g. other agents and webservers) on behalf of the participants who have delegated tasks to them. For mediating agents a (preference) model of these participants is indispensable. In this paper, a generic mediating agent architecture is introduced. Furthermore, we discuss our view of user preference modelling and its need in agentmediated electronic commerce. We survey the state of the art in the field of preference modelling and suggest that preferences of the involved participants can be modelled by learning from their behaviour. In particular, we employ an existing machine learning method called inductive logic programming (ILP). We argue that this method can be used by mediating agents to detect regularities in the behaviour of the involved participants and induce hypotheses about their preferences automatically and dynamically. Finally, we discuss some advantages and disadvantages of using inductive logic programming as a method for preference modelling and compare it with other approaches.

Learning weak constraints in answer set programming

“…The task is then to find a hypothesis that, together with B, covers all the positive examples but none of the negative examples. While in previous work ILP systems such as TILDE (Blockeel and De Raedt 1998) and Aleph (Srinivasan 2001) have been applied to preference learning (Dastani et al 2001;Horváth 2012), this has addressed learning ratings, such as good , poor and bad , rather than rankings over the examples. Ratings are not expressive enough if we want to find an optimal solution as we may rate many objects as good when some are better than others.…”

Warning system for online market research – Identifying critical situations in online opinion formation

“…For example, Choudhury [9] analyzes past activities of online groups in order to predict future activities. Dastani et al [13] predict user preferences based on their e-commerce activities. Choudhury et al [8] propose a method for forecasting the flow of communication in online communities.…”

On agent-mediated electronic commerce

“…5 To do this, the notification agent needs to have a profile for the user. This profile can be obtained in many different ways: through observing the user's behavior [15], through direct elicitation techniques [114], or through inductive logic programming techniques [28]. Once the profile is installed in the agent, it can notify the user whenever an appropriate good/service becomes available (i.e., the user's profile matches a good/ service catalog).…”

“…27 The semantic Web provides a new solution for the development of Web-based intelligent agents. There are two semantic Web languages: Resource Description Framework (RDF) 28 and Darpa Agent Markup Language (DAML).…”