Data Reduction Workflow Patterns Mining and Optimization based on Ontology

Focusing on massive high-dimensional data reduction, this paper established the ontology model of data reduction, to improve the related theoretical methods of data reduction, and to propose an ontology-based data reduction system architecture. Data reduction workflow model, workflow pattern mining, and workflow optimization and data reduction experiment design based on knowledge base are studied to achieve the accumulation, sharing and reuse of data reduction knowledge and enhance the intelligence level of data reduction process as well as the credibility of reduction results. A mechanism of meta-reduction system framework and its technology is put forward so as to improve the availability, flexibility and applicability of data reduction system.