The Application Of Data Mining In The Personalized Recommendation Service Of Readers’ Reading

The service modes of academic university libraries are gradually changing to thedirection of informatization and automatization with the application of computertechnology in the libraries. And the personalized information service has become thebright point of the new modes. The researchers are studying how to achieve the servicebetter，and it has become an importance direction of the researchers.This paper uses the data mining technology to research the daily management andreaders’ reading of university libraries. The result can be used to make the correctguidance and develop the personalized recommendation for the readers. At present, mostof management information systems of university libraries have not the functions of datamining. Having analyzing the basic data in circulation system of the library, it designs theoverall structure of the personalized recommendation system and the lending process forthe readers. Having analyzed the requirements of the system, it also establishes the systemmodel. The functions of the model modules are introduced, too. It studies the dataacquisition and mining in the personalized recommendation system. Firstly, the paper doessome researches on clustering analysis technique and its application in library，has aclustering analysis on the readers and the books using clustering algorithm. It analyzes thereaders by cluster classifying, summarizes their reading actions, excavates the clusteredreadership data by association rules, and then tries to find the relevance between thereaders and books. Then the personalized recommendation service can be supplied for thereaders.It includes a series of reasonable suggestions about optimizing collection and theefficiency reader service, and lays the solid theoretical basis for the management andrealizing the personalized service in the libraries.