EXTENDED ABSTRACT: In recent years, materials science is experiencing a shift from trial-and-error to the new paradigm of data-driven, with "data + artiffcial intelligence (AI)" at its core. It is widely recognized that data is the foundation for unleash the power of AI. Traditionally, the primary function of data has been to provide facts and is used as a digital basis for scientiffc research, technology design, validation, and decision-making. While in the context of AI, data is an information carrier that reflects a comprehensive effect from the complex influence of multiple factors. By processing and mining massive data, AI methods provide a means to reveal the implicit relationship between complex parameters by establishing correlations between data. Since data’s role has fundamentally changed in the fourth paradigm, a set of new principles must be put forward to ensure that materials data is suitable to the needs of AI, or AI-ready. Therefore, it is necessary to establish a standard system, i.e. a set of standards, to regulate the data content, generation, recording, organization, and utilization of data to enable the smooth sharing of multi-source heterogeneous data. In addition, each data can be viewed from the data generation and data application. The standard subsystems for data are generated from both dimensions to serve different needs.

Keywords: Materials data standard; Standard system; AI ready