Research On The Application Of Social Network In University Library Based On Co - Word Analysis

The birth and development of social networks have brought unprecedented opportunitiesand challenges for the librarianship. It has been pursued and admired among the vast majority ofcollege students particularly due to the features of wide coverage, fast transmission, highengagement and high viscosity. Obviously social networks have become a part of life inuniversities. In a manner of speaking, the widely use of social networks provides a chance and aplatform which is easy to combine online readers with university libraries. So libraries inuniversities can set up a deeper connection to users while communicating with readers, whichmeans university libraries must take the initiative to offer services for lessons and scientificresearch in universities, and their functions such as education and transmission of informationmust be exerted.Based on the domestic and international researches, this article strives to propose rationalsuggestions on how university libraries should fully explore their roles to make full use of tsocial network to serve users in the constantly-changing environment. We hope to distinguish ourresearch from others that based on website investigations, thoroughly probe into potentialinformation implied in the contents of the literature. Hence, on the basis of studies on theexperience of other scholars, we used co-word analysis to extract characteristics from articlesabout how university libraries make use of the social network via sparse principal componentanalysis method and penalized matrix decomposition, and we did cluster analysis for the sake ofrevealing hotspots on this theme. Thereby, we introduced some references for social networkingapplications in the university library approaching. This artcle, through those following chapters,discussed some problems:Chapter one discusses the background and significance of this study, reviews the researchstatus about social networking applied in university libraries both domestically andinternationally.Chapter two describs relevant concepts on the social network, meanwhile takes examples torepresent the current state in which university libraries used the social network.In chapter three, we summarize the co-word analysis theory and general steps, then somedisadvantages existing in the co-word analysis are pointed out. We put forward improvementmethods as well.Chapter four, we surveys literature on university libraries making use of the domestic socialnetwork, by counting up the number of characteristics in this field. Then a co-word matrix ofthose characteristics is constructed. We extract principal characteristics via a method called sparse principal component analysis, which is enough to reveal the research highlights aboutuniversity libraries using social networks.In chapter5, we proposed a new method named penalized matrix decomposition to extractcore characteristics in texts. Then, we did an experiment on the co-word matrix which selectedthe same topic on university libraries using social network. After using PMD via MATLABR2012a to reduce the dimensionality of the co-word matrix, we compare the core characteristicswhich are extracted by PMD with those extracted by principal component analysis. Accordingly,we did cluster analysis.Chapter6, tables a better proposal about how to improve university libraries using SNS sothat library resources can be utilized to the maximum. In the last chapter, we summarize thisthesis and point out further research.