Cover Image

AI-Powered Innovation in Materials Science The Role of Language Models in Discovery and Design.

Accelerate materials innovation using language models and machine learning methods Language models and machine learning are transforming how researchers discover, design, and optimize advanced materials.

Bibliographic Details
Main Author: Jiang, Xue
Other Authors: Su, Yanjing
Format: eBook
Language:English
Published: Newark : John Wiley & Sons, Incorporated, 2026.
Subjects:
Materials > Technological innovations.
Artificial intelligence > Engineering applications.
Electronic books.
Online Access:Connect to the full text of this electronic book
Table of Contents:
  • Cover
  • Title Page
  • Copyright
  • Contents
  • Preface
  • Chapter 1: The Revolution of AI for Materials
  • 1.1 Introduction
  • 1.2 What Is AI4Mater?
  • 1.2.1 Definition
  • 1.2.2 History
  • 1.2.3 Motivation
  • 1.3 Foundations and Frontiers
  • 1.4 Previous Works
  • 1.4.1 Materials Data Infrastructure
  • 1.4.2 Machine Learning in Materials
  • 1.4.3 Autonomous Experiments
  • 1.4.4 Intelligent Computation
  • 1.4.5 Intelligent Manufacture
  • References
  • Chapter 2: Fundamentals of Language Models and NLP
  • 2.1 Introduction
  • 2.2 Historical Evolution of NLP
  • 2.2.1 Statistical Language Models
  • 2.2.2 Machine Learning Models
  • 2.2.3 Deep Learning Models
  • 2.2.4 Pre-training and LLMs
  • 2.3 Core Architectures in Modern NLP
  • 2.3.1 Language Models
  • 2.3.2 Encoder-decoder
  • 2.3.3 Transformers
  • 2.4 Training and Optimization Methods for LLMs
  • 2.4.1 Pretraining Strategy
  • 2.4.2 Fine-tuning Strategy
  • 2.4.3 RAG
  • 2.4.4 Agent
  • 2.4.5 Reinforcement Learning
  • 2.4.6 Corpus Building
  • 2.5 Major Language Model Families
  • 2.5.1 Word2Vector
  • 2.5.2 BERT
  • 2.5.3 GPT
  • 2.5.4 T5
  • 2.6 Practical Tools and Libraries for NLP
  • 2.6.1 Hugging Face
  • 2.6.2 PyTorch
  • 2.6.3 NLTK
  • 4.2.1 Foundations of Unsupervised Word Embeddings for Materials Science
  • 4.2.2 Encoding Scientific Knowledge Through Semantic Relationships
  • 4.2.3 Element Embedding Spaces and Periodic Table Correlations
  • 4.2.4 Discovering and Predicting Materials with Word Embeddings
  • 4.2.5 Historical Validation and Temporal Trends in Discovery Prediction
  • 4.2.6 Unconventional Discoveries and Knowledge Beyond Composition
  • 4.2.7 Conclusion
  • 4.3 Context Similarity for Designing UHEAs
  • 4.3.1 From Linguistics to Materials Science
  • 4.3.2 Element Embeddings and Chemical Intuition
  • 4.3.3 From Similarity Scores to Alloy Discovery
  • 4.3.4 Correlation with Thermodynamic Predictors
  • 4.3.5 Designing Lightweight HEAs
  • 4.3.6 Integration with ICME and KGs
  • 4.4 Conclusion
  • References
  • Chapter 5: Materials Transformer-based Models
  • 5.1 Introduction
  • 5.2 Encoder-based Models for Materials
  • 5.2.1 BatteryBERT
  • 5.2.2 MatSciBERT
  • 5.3 Decoder-based Models for Materials
  • 5.3.1 Chemistry Assistant: A Decoder-based Model for MOF Synthesis Text Mining and Prediction
  • 5.3.2 NatureLM: Unlocking the Language of Nature for Scientific Discovery
  • 5.4 Conclusion