作　　者： ; ; ; ;

机构地区： 中南大学信息科学与工程学院

出　　处： 《计算机研究与发展》 2008年第8期1371-1378,共8页

摘　　要： 随着Web技术及其应用的快速发展,XML已经成为互联网上信息表示和数据交换的一个重要标准,其作用已深入到网络社区的每个角落;针对XML文档进行群体搜索的特点与不足,提出利用群智能算法的概率变换规则对其进行改进,首先采用路径离散化规则,结合XML半结构化的特点及概率知识,再融合粒子群算法与蚁群算法进行动态群体搜索,而群体自适应杂交、多次编码、迭代选择等不仅可以提高数据搜索的范围、精度和收敛的效率,而且可以避免早熟,降低算法的复杂度.仿真实验表明这种融合方法具有更好的查询效果. With the fast development of Web technology and its application, XML has become the important standard of information expression and data exchange on the Internet, XML functions have reached every corner of the network community. Considering the characteristics of XML for multi- objective optimization and the shortcomings of XML query, an optimization method using probabilistic rules of swarm intelligence algorithm is proposed. It adopts a path scattered rule, and combines PSO （particle swarm optimization） with ACO（ant colony optimization） to conduct dynamic swarm query by XML characteristics of semi-structured and the probabilistic rules to improve XML query, especially the PSO has the fast stochastic overall search ability, but it is unable to use the feedback information. The swarm actions are taken in turn towards the targets： swarm self-adaptive cross, encoding repeatedly, iterative choice, etc. , which leads to the following good results, widening data search range, improving search precision and convergence efficiency, avoiding premature convergence, and reducing complexity of the algorithm. The simulation experiments show that the proposed combined method has a preferable query effectively.

关 键 词： 概率查询 路径离散化 粒子群算法 蚁群算法 群体

领　　域： [自动化与计算机技术] [自动化与计算机技术]