Deep Web查询优化算法研究  

Research on Optimization Algorithm for Deep Web Query

作　　者：张亮 陆余良[2] 袁桓 张旻[4] 

机构地区：[1]解放军69012部队 [2]解放军电子工程学院网络工程系 [3]解放军96634部队 [4]解放军电子工程学院信息工程系

出　　处：《小型微型计算机系统》2012年第3期552-557,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(60972161)资助;国家242信息安全计划项目(2005C62)资助

摘　　要：Deep Web查询是在指分析接口属性及其丰富的语义信息后构造的用于向数据源请求特定数据的语句,其质量将影响查询结果相关度的高低和查询代价的大小.为优化查询,提出一种基于量子遗传算法的优化算法,以Deep Web查询的实数二进制串为输入进行量子编码,引入了球面解空间多子群并行寻优机制、群间染色体置换操作和量子变异算子以丰富种群多样性、提高算法的寻优效率.实验结果表明,该算法在R-Precision、覆盖率上具有一定的优势,能够有效地减少查询次数.Deep Web query is a request sent to data sources after analyzing the properties and plentiful semantic information of interfaces to get the particular data,qualities of which can affect the relativity of results and the query cost.To optimize the query,this paper proposes an optimizing algorithm based on quantum genetic algorithm,which takes the binary string of Deep Web query as input that is encoded in quantum state.The algorithm introduces multi-swarms parallel searching mechanism in sphere solution spaces with a permutation operator among chromosomes and a quantum mutation operator to enrich the population and promote the efficiency of searching optimal solution.The experiment results indicate that comparing with other genetic algorithms,the algorithm performs well in R-Precision and coverage,and can reduce query count effectively.

关 键 词：DEEP WEB 查询优化 量子计算 量子遗传算法 并行寻优 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]