Design And Research Of The Library System Of Personalized Service Based On Data-mining

Entered the compound stage of the traditional and digital library, libraryservicesmove towards the information technology, digitized direction, personalized service isthe main features of the digital library. At present the vast majority of our libraries,especially collegelibrary digital, most of them are to invest a lump sum funding, allor part of a collection of documents digitized and made into a library managementsystem to achieve the online book search, the reader make an appointment andrenewal, and other functions, these functionschange the traditional service modenetwork service, these services are passive to provide readers with books andmaterials, borrow the internal data link can not be insight, you can not find the needsof readers take the initiative to push information to achieve a personalized service.In this context, data mining technology for the library loan data.Knowledge ofconcepts, types of algorithm, the task of data mining and association rules analysisof the characteristics of the library loan data, compressed support a transaction-basedcorrelation algorithms applied to the library to borrow on this basis informationmining, the algorithm by compressed scanning the transaction database to improvethe efficiency of mining and support settings mining more useful rules. Furthermore,on the three major recommendation algorithm compared to borrow the record bookcirculation time is short, the recommendation based on association rules can not berecommended. Library loan data is generally not involved in the ratings of thereaders of books, books of the features are easily computer automatically taken out.Various factors to achieve the recommended method based on the content of thebook recommended. The basic idea of content-based system first analyzes the userhas to borrow books in common (type style), and then recommend the other book ofhigh similarity with those users interested in the books.Finally, ASP+SQL Server2005to build a small library of personalized service.The system functions of two parts: administrator background offline mining, usersonline recommendation. Administrator to import a certain period of time to borrowdata mining, rules stored in the rule base for recommended use. Online user loginsystem, informed by new book, the borrow Top surfing recommendation cancustomize and borrow books recommended.