位置:成果数据库 > 期刊 > 期刊详情页
基于BPSO的web服务推荐策略
  • 期刊名称:深圳大学学报(理工版)
  • 时间:0
  • 页码:49-55
  • 分类:TP393[自动化与计算机技术—计算机应用技术;自动化与计算机技术—计算机科学与技术]
  • 作者机构:[1]北京航空航天大学经济管理学院,北京100191, [2]北京林业大学经济管理学院,北京100191
  • 相关基金:国家自然科学星金资助项目(70971005,90924020);国家科技支撑计划再大专项基金资助项目(2006BAK04A23)
  • 相关项目:基于贝叶斯本体的Web知识发现研究
中文摘要:

为解决web服务的优化选择,提出一种基于离散二进制粒子群算法(binary particle swarm optimization,BPSO)的web服务推荐策略.用数学方法阐述基于服务质量(quality of service,QoS)的业务组合,将业务单元组合转换到服务组合,给出不同服务组合模式下的QoS属性值计算公式,提出web服务集和嵌套概念,对具有嵌套模式的服务组合进行逐一遍历.将基于QoS的web服务组合优化问题看成是多目标优化决策问题,提出基于BPSO的web服务组合优化数学模型,利用目标加权法简化多目标决策问题.对BPSO进行改进,构建了基于BPSO的web服务推荐仿真系统,仿真表明,该方法高效可行.

英文摘要:

To optimize the web services selection process, a method based on binary particle swarm optimization (BPSO) for recommending web services was proposed. The process composition, based on quality of service (QoS), was mathematically described and transformed into service composition. The formulas for calculating the QoS attributes in different service compositions were provided. Each service composition in the nested format was traversed by introducing the web services sets and nested formats. Treated as a multi-objective optimization decision problem, QoS-based web services selection was simplified by weighted summation. Finally, the web services recommendation simulation system was developed based on the improved BPSO. Three experiments demonstrate the proposed method is feasible and effective.

同期刊论文项目
期刊论文 35 会议论文 9
同项目期刊论文