An Integrated NLP-ML Framework for Property Prediction and Design of Steels.

Area of Science:

• Materials Science
• Computational Materials Science
• Data Science

Background:

• Steel development relies heavily on experimental processes.
• Predicting steel properties from composition and processing is complex.
• Accelerating steel discovery requires advanced computational tools.

Purpose of the Study:

• To develop a data-driven framework for accelerated steel property prediction.
• To integrate natural language processing (NLP) and machine learning (ML) for steel analysis.
• To facilitate efficient exploration and design of novel steel materials.

Main Methods:

• Utilized unsupervised ML (clustering) with NLP for steel process classification.
• Employed supervised regression models for predicting mechanical properties (yield and ultimate tensile strength).
• Developed a cloud-based graphical user interface (GUI) for user interaction and predictive insights.

Main Results:

• Achieved high predictive accuracy with R² > 0.85 for mechanical properties.
• Obtained mean absolute errors below 15 MPa for yield and ultimate tensile strength predictions.
• Demonstrated the framework's capability to provide rapid insights into steel performance.

Conclusions:

• The data-driven framework significantly accelerates steel property prediction and material discovery.
• The approach supports circular economy principles by minimizing trial-and-error experimentation.
• This work promotes sustainable innovation in steel design and development.