（1. Collaborative Innovation Center of Steel Technology, University of Science and Technology Beijing, Beijing 100083, China) （2. Kennametal Inc.，Latrobe, PA 15650, USA) （3. Institute for Advanced Materials and Technology, University of Science and Technology Beijing, Beijing 100083, China)
The material data are one of the three key tools in materials genome initiative (MGI), which have been attracting great attention worldwide. Projects on large scale databases construction and data mining have been implemented in US, Japan and other countries. The accuracy and integrity of the materials data are the foundation of data analysis and mining and they will directly influence the quality of database construction and deep extraction of the data value. The main features of materials data are high dimensions of materials attributes and complex interactive relationships. It’s worth noting that both the data mining should be associated with domain knowledge of materials and the typical requirement of materials on the outlier analysis. Education on materials data related disciplines, especially the college education on math and IT technology, will be the basic guarantee for the data being as the paradigm of innovation. The problems to be settled concerning the long term development of materials data were discussed in this paper.