The Research Of Data Refining Architecture In Data Resources

Data resources, containing a great deal of valuable information, are financial abundance accumulated in daily life. In recent years, with the capacity of data resources increasingly, the bottleneck is met for the traditional methods of data analysis and processing on the large amount of data. In this paper, a new idea is attempted by the multidisciplinary consortium for data analysis and processing.Data resources come in various properties. After the comparative research, it is found that the properties of data resource are largely similar to the properties of petroleum, and the processing of data resources such as, classification, clustering, optimization and so on, has higher goodness of fit on petroleum refining process. Referring to the nature in petroleum chemical industry, the concept and calculation on a series of properties of data resources, including data viscosity, data relative molecular mass, data boiling range, data freezing point, and data ash, are extended to be defined from data analysis and processing. According to the description of petroleum refining process, data refining architecture is proposed. Data refining architecture not only includes the process of data refining, but also achieves the data processing from the perspective of data fractionation and optimizes the data processing from the views of data refining technology.Aimed at data resources about anonymous assessment on Shanghai postgraduate dissertation, data refining process is done with different methods of three branches around data refining architecture, and experimental verification is made by the analysis of submissions, evaluations and professional descriptions. It is strongly supported the method of optimization and manipulation of data refining architecture.As the preliminary study on data refining architecture, the pioneering exploration of data resources from various views is made, and the important theoretical value and meaning is reflected. The practical data application can fully explain the scientific and validity of data refining architecture, and establish a strong foundation for the further study.