摘要
为了解决现有辩论协商模型过于抽象的问题,提出了一种改进的辩论协商架构.在构架中定义了辩论协商模型的对手、自身和环境等模型,其中对手模型是以协商伙伴为对手,自身模型的内容包括信念、愿望、意图、信任和偏爱,环境模型主要由Agent通讯语言和内容语言、协商协议和信息库组成.同时,提出了相应的算法,它通过合理地组合模型,调整论证评估器的位置,建立协议推理、论证评估、论证产生和论证选取等过程,使得协商信息充分交互,协商过程更为有效.实验结果表明,所提模型和算法是有效、可行的,算法的复杂度与协商对话的轮回有关,若轮回数为n,则复杂度为O(n).
Aiming at dealing with the problem that existing models of argumentation-based negotiation is too abstractive, an improved framework of argumentation-based negotiation was presented, in which the opponent model, self-model and environment model were defined in detail. In the opponent model, the negotiating partner was considered as an opponent. The self-model included belief, desire, intention, trust and preference. The environment model was composed of agent communication language and domain language, negotiation protocol, and information base. Meanwhile, the corresponding algorithm was also proposed. Through adjusting the location of argument evaluator and creating the process of protocol reasoning, evaluation demonstration, argument generation and argument selection, the algorithm made the exchange of negotiated information sufficiently and the negotiating procedure more effectively. Experimental results show that the proposed model and algorithm are valid and feasible. The complexity of the algorithm depends on the go-a-round of the negotiation. If the number of the go-a-round in negotiating is n, the complexity of algorithm is O（n）.
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2006年第2期129-132,156,共5页
Journal of Xi'an Jiaotong University
基金
国家重点基础研究发展规划资助项目(2004CB719401)
关键词
辩论协商
论证
协议推理
argumentation negotiation
argument
protocol reasoning
