Font Size: a A A

Web Data Mining Based On Social Network Analysis

Posted on:2013-01-16Degree:MasterType:Thesis
Country:ChinaCandidate:Y GaoFull Text:PDF
GTID:2248330371985356Subject:Software engineering
Abstract/Summary:PDF Full Text Request
In this information age, the number of web pages on the Internet is huge and they areincreasing rapidly. If you want to get what we need information on the network, the searchengine can help us get the information needed, but most of the information search is not reallylooking for, but also the accuracy of the information provided by the identification of searchengine degree. In this situation, the best way to get information is from authoritative webpages. When users use search engine to search information on the Internet, authoritative webpages should be shown to improve efficiency and quality of search engine results.This dissertation investigates web data mining based on Social Network Analysis methodsincluding research on related technologies and a study of authoritative web page discoveryfrom the web resources.The aim of the dissertation is to discover authoritative web pages from web resources.This dissertation is supposed to help people find authoritative web pages easily, and let peopleaccess useful information efficiently through applying Social Network Analysis methods toanalyze the relationship between related web pages.Several related technologies are researched in this dissertation. They are Data Mining,Data Mining technologies, the characteristics of Web data, Web Mining, Web Search Engines,Google, authoritative pages, Social Network Analysis, Degree Centrality and Social NetworkAnalysis software UCINET6. The aim of these studies is to provide the theoretical basis forthe experiment.The main method used in the dissertation’s experiment is Degree Centrality, which isfrom Social Network Analysis and is used to analyze the relationships between web pages.Authoritative pages can be considered as web pages which have been referenced by a lot ofother web pages. This means that they are reliable and have high-level acceptance. Severalauthoritative web pages were found through calculating their degree by UCINET6in theexperiment presented in this dissertation. The study showed that the main thesis of thisdissertation is reasonable, and appropriate authoritative web pages can be found through themethod used in the dissertation. This work can be extended further in the future, such as bycalculating relativity, eliminating repetition and extending the dataset. These will bementioned in the future work of this dissertation.
Keywords/Search Tags:Data Mining, Web Search Engines, Authoritative Page, Social Network Analysis, DegreeCentrality, UCINET6
PDF Full Text Request
Related items