Research of Hadoop-based digital library data service system

In view of the rapid growth of digital library data, the data of different structures are increasing day by day. The traditional data processing model can not be qualified for the storage and management of massive heterogeneous data. A Hadoop-based digital library service system is designed and implemented for all the practical problems in the big data field. The system makes full use of the massive storage capacity and strong parallel computing power of Hadoop platform. It designs a custom business process model and extended component interface, to achieve distributed storage and computing of digital library system. In order to further verify the efficiency of the system, respectively, through the Hadoop cluster and stand-alone system, the different sizes of digital library resources are experimented for analysis. Experiments show that the Hadoop-based data processing system can significantly improve the data processing capacity, and be more suitable for mass unstructured data storage, retrieval and analysis.