Page 13 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2018 Vol. 42

012 Journal of Library Science in China, Vol. 8, 2016

analysis of information, The clustering of user needs can help us compare and analyze similar
demands, help users optimize their demands, and promote the expression of demands. The
clustering of user behaviors can find the law of user behaviors of resource usage, and provide data
support for user personalized recommendation service.
(3) Information analysis techniques
Digital libraries of the big data era should make full use of information analysis techniques,
incorporate big data analysis technology into digital libraries, regard data analysis as a development
work of digital libraries, and take the analysis results as the high-end products provided to users.
Among these analysis techniques, literature information compiling technology can provide
support for tracking of cutting-edge research; On-line Analytical Processing (OLAP) can online
analyze users’ behaviors of digital library usage, and provide data support for the management and
operation of digital libraries; other quantitative and qualitative analysis tools and techniques, such
as Hadoop, SPSS, CiteSpace, etc. should also be fully utilized by digital libraries.
(4) Retrieval technology
The retrieval technology deployed by next generation digital libraries should not be constrained
to index to a single digital library, but realize seamless cross-platform and cross-resource
retrieval using network retrieval technology. This requires a retrieval network between the digital
libraries. Each digital library can achieve internal and external resource retrieval and provide
network resources. In the meantime, academic search should become a product offered by digital
libraries in universities and research institutes, in order to promote the academic research of their
staffs. Due to the self-service pattern of the usage and retrieval of digital libraries, the retrieval
interface should be made user-friendly with more powerful functions, such as semantic retrieval,
conversational language retrieval and cross-language retrieval which can automatically understand
or translate multiple languages.

2.3 Product and service mode of digital libraries

Literature information cannot completely meet the demands of users in the big data era. Users
now need more diversified information. In addition to officially published literature information,
they also need videos, images, unofficially published information and network information, even
regenerated information.
(1) Digital library service products
Academic resource constructors are providing more types of products, which almost include all
of the digital library resources and are constantly expanding, making a bigger impact on digital
libraries. Therefore, next generation digital libraries must offer more diversified products to be
more competitive. Besides literature and network information services, digital libraries should be
able to provide renewable information services, such as knowledge base, method library, inference