Research On Social Tags-classification Integrated Navigation And Its Applications

With the popularization and development of the Internet, the contradictionbetween information expansion and low information retrieval efficiency becomesmore prominent. The research on how to improve the accuracy and efficiency ofinformation retrieval has become one with important theoretical and practicalsignificance. Combined with the National Natural Science Fund Project "Researchon the combination of the speed of the knowledge extension optimization andsharing theory in knowledge super network environment"(Project NO:71071144)and the Zhejiang Province Natural Science Fund Project "Research on extensibleand adaptive knowledge service for complex product design"(projectNO:Z6110334), this paper makes deep research on the social tags-classificationintegrated navigation. Mainly research work and contributions are as follows:Firstly, after the introduction of the information classification and the social tagand the analysis of the integration methods research status, the socialtags-classification integrated navigation (SCIN) is put forward. The characteristicsof SCIN are analyzed. Based on this, combined with the navigation constructionrequirement, the building process and key technologies of SCIN are put forward.Secondly, the construction of classification and tag cloud model and theprocess and algorithm of integration are detailedly introduced. According to therelationship between the tag cloud and the classification, three integrating modes ofSCIN have been proposed, which are homepage-based, classification-based,information-based. To integrate social tags and classification, H-ATR algorithm, C-ATR algorithm and I-ATR algorithm are put forward as the difference of the tag range.Thirdly, to reorder retrieval result and actualize information retrieval based onSCIN, the topic similarity and time weighted page rank algorithm (STPR) isproposed based on the information recommendation by Web information link in theinformation retrieval system.At last, the function model and system framework of enterprise informationretrieval system based on SCIN are constructed. The system is developed and put into practice successfully. Based on this, to analyse and evaluate the effect of SCINbased enterprise information retrieval system, a users’ information behavior basedevaluation method (WP@N) is used.