摘要
基于遗传算法的并发谈判在电子商务应用中具有独特的优势,但已有的研究尚未考虑议题之间的相关性、动态权重的变化规则依赖于对手信息的获取,极大地限制了其使用价值.对此,提出议题分组的方法以解决议题的相关性问题,而议题权重的变化则采取从历史资源中发掘知识的方法进行动态调整.构建了基于遗传算法的关联性议题并发谈判模型,给出了模型的形式化描述、谈判算法设计和动态权重更新方案.通过对模型的实验和比较分析,证实了该方案能够更方便地满足用户谈判的多样性需求,解决谈判中议题关联性的问题,而且能够快速、有效地得出最优谈判结果.
The concurrent negotiation based on genetic algorithm has special advantages in e-commerce applications. But, the relativity of issues and the dynamic weights of issues are not taken into account in existing research. Thus, this paper proposes a solution to these problems by grouping the issues and adapting the dynamic weights, according to data mining from history resources. The concurrent negotiation model of relative issues is described in detail, including the formal definition of the model, the design of concurrent negotiation algorithm, and the update scheme of dynamic weights. The experimental results show that the model can meet the requirements of different negotiations, solve the issues’ relativity problem in the process of negotiation, and improve the negotiation efficiency.
出处
《软件学报》
EI
CSCD
北大核心
2012年第11期2987-2999,共13页
Journal of Software
基金
国家自然科学基金(61272406)
关键词
遗传算法
并发谈判
关联性议题
动态权重
genetic algorithm concurrent negotiation relative issue dynamic weight
