Scalable Databases and Modelling Systems for materials intelligent design
Brief Introduction
SDaMS is an integrated software platform developed by MICID group at Beihang University since 2022. It is dedicated to integrate the materials database & analysis, data mining & machine learning, modelling & knowledge discovery. The platform provides simple graphical user interface to database. It supports fast for both local and remote database construction and editing based on SQLite and MySQL, as well as the construction and application of robust physical and mathematic models through various machine learning algorithms. It provides users with an integrated workflow platform, including database construction, data analysis & mining, machine learning model construction, and visual editing. Additionally, the research group has collected and built a series of databases and model libraries to complement the use of SDaMS.
Database Management
SDaMS provides a simple and convenient database management tool. It can automatically generate local or remote databases based on UI files created by users. In terms of data import, it not only offers visual editing of data entries but also supports batch import or modification of large amounts of data through JSON or Excel files. It even enables seamless integration with SPaMD software to import calculation results with just a click. SDaMS simplifies the process of database editing, providing users with an efficient and user-friendly solution.
Data Analysis & Mining
SDaMS provides a convenient and user-friendly big data analysis platform, which includes features such as multi-criteria data retrieval, specified data export, scatter plots, line charts, pie charts, and other common data statistics. Additionally, it encompasses commonly used analysis methods in materials science, such as automatic analysis of crystal symmetry. With SDaMS, users can easily perform various data analysis tasks, enabling efficient exploration and visualization of data in a range of scientific disciplines.
Intelligent Design
SDaMS provides various descriptor generation methods and machine learning methods, allowing seamless integration with databases for convenient and efficient model construction. Additionally, SDaMS comes with a diverse range of built-in databases, including specific databases, such as Superhard materials, Biocompitable materials, Low-dimensional materials; ICSD-derived prototypes, PCD-derived prototypes, ASM-derived prototypes, which further enhance its capabilities. Furthermore, SDaMS incorporates a model library developed by our research group, offering a convenient and efficient means to predict material structures and properties, promoting state-of-art material research & development.
Features
- A user-friendly GUI for operation and visualization of database and modelbase;
- Local and remote databases are supported through interface file
- Database construction; New database by interface file, copy of current interface
- Datasets operation; Add, delete, revise, search
- Designed GUI for Database; Superhard materials, Biocompitable materials, Low-dimensional materials; ICSD-derived prototypes, PCD-derived prototypes, ASM-derived prototypes, etc.
- Algorithms for machine learning includes various supervised ML models, unsupervised ML models and feature engineering methods.
- Modelbase includes Miedema Calculator (ML version), Pourbaix calculator, EPCK calculator, Enthalpy calculator etc.
- Visualization tools include the typical drawing of curves such as scatter, lines, pie and collumns.
Gallery
Updates
- Updated in Ver.1.0: Database construction and statistical analysis, supervised and unsupervised machine learning methods, various mathematic and physical models.
License
License statement：The SDaMS is currently copyrighted and distributed free of charge for academic, scientific, and educational, and non-commercial users with our permission. Commercial users may also use this software with no cost until a license is established, or with a contract for payment for the technical support. This software is provided ‘as it is’ without any expressed or implied warranty. To obtain the SDaMS, one can send an email with the registration form to the email: zrfcms@buaa.edu.cn