Application Of Technology In The Library Of Personalized Information Service Data Mining

In the progressive modern society (social division of labor) for its sophistication and complexity, it has been evidenced that, in the library system, the general public possesses an ever-increasing demand for a high level of information accuracy, timeliness and personalization. The Library, as an organization to collect and store documentation in various formats, has expanded its overall capacity for the huge amount of information growing daily. Consequently, the general public faces a serious challenge of identifying the appropriate information quickly and accurately from the information explosion.This article analyzes the existing library information management system (LIBMS) and indentified that the demand for the individualized information service has not yet been met. Furthermore, this article investigates the weakness of the current LIBMS and suggests relevant enhancement for the LIBMS, particularly a sub-system for personalization has been developed. The focuses of this article are the applicable technology of cluster analysis and discovery of association rules and its respective applications in the data-mining discipline.The following summary illustrates the above focuses:(1) Requirement analysis and business process analysis for developmentThis article develops a design for the overall structure and function module. Various system analysis methodologies and tools have been adopted, including Unified Modeling Language (UML), particularly case diagram, activity diagram, class diagram, sequence diagram etc.(2) A data cleaning process has been conducted for the raw data in preparing the master datasets. A sample analysis established in this article to select the association rule mining support number. These support number is then verified through the K-means via the weka tool(3) This article also proposes the SQL-based association rule mining method.This proposal is based on the following critical reasons:a. The book association rule mining just mining frequent2-itemsets can be enough.b. the classic Apriori algorithm processes,-the direct use of the DBMS mining association rules.During the process of the formation of candidate2-itemsets,the algorithm can count out the support number of this candidate2-itemsets.There is no need to scan the database successively to complete statistics of support number, consequently greatly enhancing the efficiency of the implementation of the algorithm.At last, Library association rules mining has been achieved and mining results are applied to personalized information services.(4) The development environment integrates the SSH framework with the adoption of the MVC methodology. This method is demonstrated via the example of the "function realization of resource retrieval", i.e. how to achieve all layers in order to realize the B/S structure in the proposed Library individualized information subsystem.Finally, comprehensive summary of research and further Outlook for the future research directions are given in this article.