Page 12 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2018 Vol. 42

SU Xinning / Opportunities and challenges faced by digital libraries in the big data era 011

with other resources like literature resources and target resources, it will become a highly efficient
resource of knowledge service.

2.2 Application of digital library technology

In the big data era, digital library technology has migrated from local data processing to data
processing of a wider field. From the perspective of the technology system of digital libraries, it
includes data acquisition, data processing, architecture construction, knowledge mining, analysis,
prediction, result presentation, service technology, etc. How to apply and incorporate big data
related technology into digital library field, is now a problem faced by the digital library field.
(1) Semantic technology
The big data environment needs semantic technology to help connect the large quantity of
complicated data. Digital libraries should think about how to automatically fuse the semantics
defined in dictionaries (subject headings, classification charts, etc.) into the related information of
digital literatures. Of course it is not practical to label manually the semantic relations of such a
large amount of data. With the aid of dictionaries, we must adopt artificial intelligence technology,
ontology, semantic analysis technology, automatically label the semantic relations of data and
make the data meaningful so as to promote the knowledge expansion and knowledge exploration
of the collected resources.
The application of semantic technology in the field of book intelligence is not uncommon.
Tools like book classification and Chinese Thesauri were created based on semantic relations. But
these tools only help us define the semantic relations between literatures rather than the semantic
relations among the knowledge points inside the literature or information which is required in a
big data environment. The establishment of the semantic relations among knowledge points can
facilitate users’ access to knowledge, and open a channel for them to obtain the useful knowledge
through keywords and semantic relations.
(2) Data clustering technology
Clustering is a process which puts similar and related information or data together. It is one of
the effective means of efficient utilization of multifarious and a huge amount of data. In digital
libraries, clustered information resources can play a more significant role in information service,
information analysis and information utilization. Clustering technology can not only gather
literature information resources according to certain attributes or characteristics, but also be
applied to user requirements, user search behavior processing and analysis.
In a big data environment, only a small portion of data is utilized. Clustering technology
provides a good way of full usage of this small portion of data. For instance, the clustering of data
resources can separate them into small data sets of close relations and similar themes, and these
small data sets are perfect for problem oriented demand, as well as users’ retrieval, selection and