An evolvable and transparent data as a service framework for multisource data integration and fusion

Combining data from multiple sources is a means of enabling unified and comprehensive description of objects in high-dimensional space and helping unlock the potential value of such data. In recent years, more and more studies have focused on this field of research. However, challenges posed by separately stored data and comprehension barriers about different systems hinder the integration of data from different sources. To overcome these problems, this paper proposes a Transparent Data as a Service framework, a novel approach combining Transparent Computing and Representational State Transfer (REST) Web Services based on Linked Data. This framework is capable of integrating data from different sources and offering data services in a transparent way. That is, consumers use data services without the need to know details of where or how the data are stored. Our framework is transparent on three levels: transparent data resource integration, transparent data fusion and transparent data service provision. The Data Model Pool and Data Resource Pool are able to evolve as new data models and datasets are generated in the provision of data services. Finally, we demonstrate the feasibility of the framework by implementing a prototype system.