Design And Implementation For The Application Of Digital Library Personalized Service System

Digital library not only provide useful information resources,also needs and provide personalized information service for different users,in order to enhance service quality and level of the digital library,improve the utilization rate of resource information.With the development of information technology and network technology,as well as the library digital resources become more and more rich and perfect,all this is possible.According to the demand of the digital library personalized service,this paper proposes a web-based digital library personalized service mechanism and its related algorithm.The main work is as follows:1.In this paper,we have done analysis and elaboration deeply for the Web mining technology in the field of data mining theory.Based on the analysis of the current to provide personalized service,on the basis of the connotation,characteristic and the way is pointed out that at present to solve the problems that exist in the library individualized information service.2.Based on the data in the traditional collaborative filtering algorithm is sparse and the shortage of the cold start,put forward the collaborative filtering algorithm based on user interest spread.In this algorithm not only solve the problem of user interest similarity calculation,users into account other neighbors of target user interest preference.In the case of data sparsity to ensure the effectiveness of the algorithm.The simulation experiment shows that the algorithm is feasibility and validity.3.Design and implementation of personalized recommendation system based on Web log mining prototype.Obtained by the improved Apriori algorithm for data preprocessing phase of the transaction database and generate frequent access patterns mining association rules.According to the frequent access pattern and association rules to automatically generate recommendations,online provides to the user.Based on user interest model library involved the analysis of the technical content,design based on automatic modeling technology,vector space model representation and the corresponding data mining algorithm of user interest model library.4.This paper put forward the separation design,online and offline integration to realize personalized recommendation of design thought,system is analyzed from the data preparation,data collection and the user interest model to online recommendation engine produces results set of recommendations,finally realized the function of each module design.Aiming at the current problems existing in the digital library personalized information service system,the synergistic effect of user interest spread into consideration,to adapt it to the user personalized needs,to raise the utilization ratio of library resources better.The system was tested,the test results show that the system can meet the needs of users,its overall performance to achieve the system the expected design goal.