The ScholarNet and Artificial Intelligence (AI) Supervisor in Material Science Research

Area of Science:

• Artificial intelligence
• Material science
• Robotics

Background:

• AI and robotics are increasingly used in chemistry and material science, but current applications focus on execution rather than idea generation.
• The rapid growth of scientific publications leads to overpublishing, making it difficult for researchers to identify novel interdisciplinary research opportunities.
• Innovation often arises from interdisciplinary research, yet identifying these opportunities is challenging due to information overload.

Purpose of the Study:

• To develop an AI supervisor that assists in the crucial phase of research idea generation and novelty assessment.
• To address the challenge of identifying novel, interdisciplinary research directions in the face of extensive scientific literature.
• To provide researchers with a tool that supports the discovery of new ideas and potential breakthroughs.

Main Methods:

• A deep learning-based AI supervisor was developed.
• The AI supervisor was trained on correlation-based ScholarNet data from scientific publications.
• The system is primarily tailored for material science applications.

Main Results:

• The AI supervisor effectively recommends research ideas relevant to material science.
• It analyzes the novelty of suggested research ideas.
• The system provides comprehensive guidance to researchers, supporting both idea generation and novelty assessment.

Conclusions:

• The developed AI supervisor serves as a valuable digital infrastructure for material science research.
• It enhances the process of scientific discovery by supporting idea generation and novelty assessment.
• This AI tool has the potential to foster innovation by identifying interdisciplinary research opportunities.