|
Digital library can provide readers with hundreds of kinds of database information, for the reader information retrieval provides great convenience. However, the digital library for the data collection and processing lack of systematic norms and principles, resulting in the form of database, data preservation methods, forms of data query, increase the difficulty of data query, plus the digital library of different database systems can not form mutually compatible search methods, in order to obtain the required data have to change frequently. If you log in different database systems at the same time, not only affects the effect of literature retrieval, but also increases the error rate of literature retrieval methods. In order to process the literature data of digital resources with ever-expanding data capacity, this paper pushes information resources precisely for users’ reading, lets the system independently classify users and identify information according to different characteristics of users, and then makes personalized recommendation according to users’ needs. Comprehensive use of user characteristics and collaborative filtering technology, based on users’ current reading preferences as the basis for the construction of library accurate recommendation service model for users’ reading preferences, to predict users’ future reading preferences to meet the personalized needs of users. The performance of the model is verified by experiments.
|
