Web信息颗粒关联强度研究  

Association Strength among Web Information Granules

作　　者：刘平峰[1,2] 杨柳[1,2] 姚梦迪[1,2] 吴钟[3] 

机构地区：[1]武汉理工大学经济学院,湖北武汉430070 [2]武汉理工大学电子商务与智能服务研究中心,湖北武汉430070 [3]武汉理工大学华夏学院,湖北武汉430223

出　　处：《武汉理工大学学报（信息与管理工程版）》2015年第4期399-402,共4页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：教育部人文社会科学基金资助项目(10YJAZH050);教育部人文社会科学研究青年基金资助项目(14YJCHZ165);湖北省自然科学基金青年基金资助项目(2014CFB353)

摘　　要：粒计算的引入提高了对非结构化海量Web信息的处理能力,但目前研究仅限于Web链接结构、粒度划分、信息颗粒树形成等方面,未涉及信息颗粒之间的关联强度分析。针对Web信息筛选的不完整问题,提出基于多层次粒度划分的Web信息颗粒关联强度算法,通过获取Web网页间直接或间接的链接关系进行矩阵运算,对同层次不同类别Web信息颗粒间的关联关系进行了定量分析,并在实验中验证了算法的正确性和有效性。解决了信息使用单一性和决策参考片面性问题,为获得全面信息和构建综合决策信息系统提供了理论参考。Granular computing can help improving the capability of handling massive unstructured Web information. Howev-er, the current studies are limited to the structure of web links, granularity division and the building of information granules tree;and the association strength among information granules were never analyzed. For the problem of incomplete Web information fil-tering, an algorithm based on multi-level granularity division was proposed to calculate the association strength among web infor-mation granules. The algorithm gave a quantitative analysis to the same level web information granules of different types by acqui-ring direct or indirect links between web pages to do matrix operation. The correctness and effectiveness of the algorithm was veri-fied by the experiment. The algorithm can solve the problems about unitary and one-sidedness effectively when using information or making a decision. It provides a theoretical reference to obtain comprehensive information and build a composite decision infor-mation system.

关 键 词：WEB信息 粒计算 WEB链接 关联强度 

分 类 号：G350[文化科学—情报学]