Web数据库大小估算方法的研究与应用-计算机应用技术专业论文.docxVIP

西安 西安理工大学硕士学位论文 Abstract Abstract II万方数据 II 万方数据 万方数据 万方数据 Title: THE RESEARCH AND APPLICATION OF THE WEB DATABASE SIZE ESTIMATION METHOD Major: Computer Application Technology Name: Lina Liu Supervisor: Prof. Quanzhu YAO Abstract Signatu With the rapid development of the network technology ，the Web turned into a vast complex data source，so the Web has become an important way of accessing to information. The Web can be divided into Surface Web and Deep Web (also known as the Hidden Web or the Invisible Web). Compared with the Surface Web，Deep Web contains richer and better structured and more focused inforτnation. In order to use huge amounts of information on the Web more efficiently and accurately，researchers have conducted a lot ofresearch in the field of Deep Web. Especially in Deep Web data integration ，has accumulated a considerable amount of research results. The focus of this thesis is a new problem of Deep Web research，how to estimate the size of a Web database. This thesis simply introduced some related work of Web database size estimation ，and pointed out their defects and deficiencies. Then aimd at the relatively perfect Web database size estimation method based on attributes correlation ，improved its defect and put forward a new method of random sampJing. The main work is as follows: (1)This thesis proposed a Web database size estimation method based on Zipf distribution. This method chose attributes that they have minimum attributes correlation to extract approximate random sample by calculating the attribute correlation. According to the approximate random sample，calculated the distribution of attribute words and the size of sample，and then on 由is basis，calculated the Zipf equation to get the frequency of attribute words，so as to estimate the size ofthe Web database. The method improved the defect of the estimation method based on attributes correlation-even selected the minimum relevant attributes to get the random sample，tw