Research And Design Of Personalized Service Module In University Digital Library

University libraries have basically realized the library automation management,and provide readers with a certain degree of personalized service,but with the development of society and people's awareness of the continuous improvement of the service,the reader has not only meet the existing books The automation system provides personalized service functions that require the system to provide better and deeper personalized services that require colleges and universities to improve existing library management systems.In order to enable readers to experience a better library personalized service,the need for large readers to borrow data through data mining technology for deep analysis.But the traditional data mining algorithm and its improved algorithm are running on a single node of the machine,and the use of serial operations,spend a long time,and can not be efficient to complete the mining task.In order to improve the efficiency of the data mining algorithm,we can consider improving the mining algorithm so that it can be processed in parallel on multiple nodes,so we can use the computing ability and storage ability of multiple nodes to complete the mining task efficiently.This paper combines the knowledge of cloud computing and data mining,and puts forward the improvement of parallelization of FPGrowth algorithm.On this basis,it designs the personalized service system of digital library in colleges and universities.The research contents mainly include the following three aspects:(1)A parallel data mining algorithm is proposed.That is,to improve the existing FP growth mining algorithm,the traditional single-node data mining algorithm FP growth algorithm improved into parallel in the cloud,the use of MapReduce algorithm to achieve friends recommended,an increase of the relevant algorithm improvements and cloud implementation.And analyzes the results of the book borrowing data mining of Guangdong Agricultural and Commercial Vocational and Technical College.(2)Design and implementation of the personalized service library system.The design and implementation of the system,including the contents include:demand analysis,feasibility analysis,functional module design,database design,interface design.(3)Using a variety of test methods to the proposed personalized service library system for testing.Performance testing aspects of the cloud server memory,CPU occupancy rate and other parameters to test;functional testing of the system to test the functional modules.