The Management Of Readers And Construction Of Personalization Service Model In Digital Library

With the rapid development of database technology and computer technology, the traditional libraries are changing to the digital library. The reader's needs of the current information of the library, has not stop at reading paper resources. Readers will be able to visit the library resources at home. they need to find the scientific and technical information and to timely, accurate, complete download it to local. At present, the library collection such as electronic journals, electronic books, and other resources data capacity has reached about 5 T, a large-scale library can achieve its twice, and keeping updating in certain rate about thousands per day. While readers can not the latest daily tracking library resources, and readers are often most in need of resources was not in time. How to make the passive library services into active service, for different readers, take different services.First of all, we focused on the individual needs of libraries, design a new set of readers management platform system, the system designed to take full account of the design physiognomy reader service objectives, the completion of the general readers information management functions on the basis of several characteristics of the particular design functional modules:The function can pride readers reminder. when the readers are going exceed the time of returning book,system will send notice to the readers, reminding readers to do "return" or "renewal" operation. So readers can avoid fines and give the reading chance to the other readers in time;Readers's borrow management, through the Contaction with readers, system can manage readers to borrow, if A has been eager to borrow book K,but book K was borrowed by B. he can make an application to the system, the system will inform B Contact A, in the end A got the book he want.this function can make the full use of resources; Readers self-service functions, readers through the Web platform can query the books he had borrowed information, such as: books count, can borrow for more, book dates, similar books (through mining algorithm); For online renewals, appointment books operation. That's can reduce the burden on library staff, and to save valuable times;Secondly, we get the information from the reader's borrow, visit the electronic files of information, download articles, documents Keywords.through cluster analysis and summed up the characteristics of the readers set. Readers cluster analysis in accordance with a certain readers to borrow books from the time of the keyword clustering, the system uses k-means clustering algorithm their calculation, the results will be stored in the cluster readers clustering results table. According to frequency, we choose the larger set of features readers to express preferences. Each time the reader to borrow acts, the system automatically and the results of its analysis of readers interested in the preservation of the table. Managers can also, if necessary, to a different time of the readers interest analysis, conducted for a class of readers interested in analysis, the results can guide the library book procurement.Thirdly, To make a similar analysis and comparison on readers and books. This article made reference by ROCK (RObust Clustering using linKs) algorithm for readers to new information and attribute similarity between the judgment. The first set of keywords identified readers that readers of all books borrowed a collection of keywords U, we can see that readers interested Set (R) is a subset of this set, each of these elements exist in the U, we have to four Keywords example marker into (1,1,1, 1), books Keywords set of elements in comparison with the U, if it exists at the U, is counted as one, does not exist for 0, can produce similar to (1,0,0,1) such a pool, set by calculating the Continental distance, with each book readers than the distance, when the distance is less than a certain threshold value, we can confirm that book is the most in line with the tastes of readers. On the threshold determined, then takes repeated training come. It is also worthy of follow-up in the work of the study.Through the above analysis of the similarity of the results, we will be able to know which resources is reader's preference. There will be information in the knowledge base, thus providing to the readers to complete the readers of personalized service. The system offers two way communication with readers, the other is the e-mail or mobile phone text message immediate way to communicate with readers; Second, the data will be integrated into the reader inquiries to use Web pages or link directly into the reader inquiries pages. When readers visit the Library Web site can be our personalized service. This system based on SQL Server 2005 database, using VS.net integrated development platform, system in the preparation process of absorption and utilization of a lot of VS.net technology.Each user has their own different backgrounds, language, culture, hobbies, and have their own type to meet the information. Therefore, the design of the Digital Library needs to be user-centric, information resources and services, various methods should be adapted to the needs of users and background, different combinations. The organization and content of services should not be stereotyped show, and to be compatible with the personality characteristics. The paper on the digital library of data mining is only a starting point, with the data mining technology and the continued progress, the rapid development of computer technology, data mining work in the library will play a bigger role.