The Research On Personalization Recommendation Technology And Its Application On Digital Library

In recent years, the rapid development of digital storage technology has made it possible to digitize almost everything. The construction of digital library, which aims at making library resources digitized, has been rapidly developed in some countries, and some research on digital library is marching into a new stage.Among many research directions of the Digital Library, the research on personalization service has become a trend. The biggest advantage of the service is that it provides individual needs of digital resources for users with different backgrounds and different intent. By analyzing users' data from web logs, personalization service identifies the user's interests, finds out the more important user's interests and the description of overall features hiding behind these users' data and then predicts the tread of their interests, which aim at providing the user with "one to one" individual information service in proper time. The work will not only improve the design of digital library's web site, but also has a great significance on the service decision of digital library.This article firstly gives an introduction about the development of domestic and international digital libraries and also an introduction about the background and the progress of China-US Million Book Digital Library Project (CADAL). Then we present an overview of personalization service technology and analyze the function and give a strategy of personalization service in digital library. Secondly, we introduce the web mining technology and give a deep analysis of the web mining process. We propose a data structure named Sequence Pattern Tree for storing web accessing information and give a Sequence Pattern Tree construction algorithm. We develop a Sequence Pattern Tree mining algorithm for discovering frequent sequence patterns on the proposed Sequence Pattern Tree. We also integrate the web mining technology and individual service in Digital Library. Lastly, we give a concrete design and implementation of personalization recommendation service system in Digital Library.