摘要
Agent negotiation has become increasingly important since the advent of electronic commerce.There are two kinds of interactions in the process of agent negotiation.One is the interaction between different agents,and the other is the interaction between the agent and the human user.In this paper,firstly,we introduce Q language,a scenario description language for designing interactions between agents and human users.Then we propose an integrating interaction framework for agent negotiation,in which both kinds of interactions are described by Q scenario.Our framework can make the interaction process open and easy to be understood by the users.Users can understand how the negotiation process goes and what is happening in the system including some erroneous or inappropriate actions caused by the negotiation agent.This gives the users a chance to terminate or change the behavior of negotiation agent in time to avoid unfavorable negotiation outcome.In addition,users can customize the agent's behaviors through visible interactions with it.
Agent negotiation has become increasingly important since the advent of electronic commerce. There are two kinds of interactions in the process of agent negotiation. One is the interaction between different agents, and the other is the interaction between the agent and the human user. In this paper, firstly, we introduce Q language, a scenario description language for designing interactions between agents and human users. Then we propose an integrating interaction framework for agent negotiation, in which both kinds of interactions are described by Q scenario. Our framework can make the interaction process open and easy to be understood by the users. Users can understand how the negotiation process goes and what is happening in the system including some erroneous or inappropriate actions caused by the negotiation agent. This gives the users a chance to terminate or change the behavior of negotiation agent in time to avoid unfavorable negotiation outcome. In addition, users can customize the agent's behaviors through visible interactions with it.
基金
This work is partially supported by the National Natural Science Foundation of China (No. 69973032)
AIEJ (Association of International Education, Japan).
