University of Science and Technology Beijing | China
Author Profile
Scopus
Google Scholar
🌟 BIOGRAPHY OF GUANGXUAN SONG
🎓 EARLY ACADEMIC PURSUITS
Guangxuan Song, born in May 1997, embarked on his academic journey with an undergraduate degree in Automation at the University of Science and Technology Beijing (USTB) from September 2016 to June 2020. His academic excellence was evident early on, receiving the National Scholarship in 2017 and being named a Beijing Outstanding Student in 2019. During his undergraduate years, he also took on leadership roles as Vice President of the Youth League Committee, Head of the News and Publicity Department, and Freshman Class Mentor.
💼 PROFESSIONAL ENDEAVORS
From September 2020 onward, Guangxuan advanced into a Ph.D. program in Control at the same university, focusing on state-of-the-art AI and data-centric projects. Alongside his academic commitments, he was actively involved in national-level research initiatives, contributing to multiple technological innovations and intellectual properties. His work has garnered six invention patents and seven software copyrights.
Guangxuan Song’s core research interest lies in Knowledge Graphs and Graph Neural Networks for scientific applications. His contributions include:
-
📌 Node Importance Evaluation in Scientific Collaboration Networks
Applied a message-passing neural network (MPNN) formulation of the PageRank algorithm with learnable parameters to rank author importance. (Published in Advanced Electronics and Computer Engineering)
-
🧪 Metal Property Prediction via Knowledge Graph and Structured Data Fusion
Built a multi-modal scientific data pipeline integrating LangChain and GPT-3.5 to predict alloy properties using a generalized Fermat point model. (Published in Corrosion Science)
-
🔍 Modeling of Noisy Real-World Scientific Data
Developed TSNet to decouple noise using multivariate Taylor expansion, improving prediction accuracy and robustness. (Under 2nd review at IEEE Transactions on Big Data)
🌐 NATIONAL AND INDUSTRIAL COLLABORATIONS
Guangxuan played a pivotal role in several high-impact national and corporate research projects:
-
National Environmental Corrosion Platform of China
-
Science and Technology Basic Resources Survey Project – “Belt and Road” Corrosion Big Data Platform
-
State Grid Corporation Project – Atmospheric Corrosion Factors and Data Mining for Power Grid Materials
In these roles, he contributed to building three major data analysis web platforms, demonstrating strong interdisciplinary collaboration and technical leadership.
🏆 HONORS AND AWARDS
Guangxuan Song has received numerous accolades, including:
-
🎓 National Scholarships (2017, 2022)
-
🏅 Dean’s Medal (2019)
-
🥇 First Prize in iCAN International Entrepreneurship Competition (2020)
-
🥇 First Prize in "Internet+" Entrepreneurship Competition (Beijing) (2021)
-
🌠 Science and Technology Star at USTB (2022)
-
🧠 Baidu Lingjing LLM Developer Recognition (2023)
-
🏃 Completion of Daxing Half Marathon (2023)
He was also recognized as an Inspirational Figure at USTB in 2020 and awarded Outstanding Graduate of Beijing Municipality in the same year.
🔗 IMPACT AND INFLUENCE
Guangxuan's work significantly enhances the scientific community's ability to extract, model, and predict information from noisy, complex, and multi-modal datasets. His innovations in graph-based ranking, uncertainty estimation, and semantic alignment contribute to cutting-edge advancements in AI for materials science and beyond. His collaborative efforts with national infrastructure platforms and state-owned enterprises emphasize his role as a bridge between academia and industry, fostering data-driven, AI-powered transformation in traditional sectors.
📚 ACADEMIC CITES
His work has been published in reputed journals, such as:
-
Advanced Electronics and Computer Engineering (ADV ELECTR COMPUT EN)
-
Corrosion Science (CORROS SCI)
-
(Pending) IEEE Transactions on Big Data
He is actively building a presence in top-tier AI and scientific publishing platforms, with a growing list of citations and international readership.
🧭 LEGACY AND FUTURE CONTRIBUTIONS
Guangxuan Song stands at the forefront of data-centric AI research, contributing to smarter, more interpretable systems in scientific and industrial domains. With his keen understanding of AI frameworks like PyTorch, Torch Geometric, and LangChain, and a hands-on approach to research and collaboration, he is poised to:
-
Pioneer real-world applications of graph-based machine learning
-
Bridge the gap between structured scientific knowledge and deep learning
-
Mentor future AI innovators through active involvement in academic leadership
💡 OTHER NOTABLE TRAITS
-
💻 Technical Skills: Expert in Python (PyTorch, Torch Geometric, LangChain), proficient in C++, embedded hardware/software development
-
📸 Interests: Passionate about photography and running
-
🤝 Soft Skills: Strong teamwork, leadership, and challenge-driven mindset
📑NOTABLE PUBLICATIONS
"Corrosion Resistant Performance Prediction in High-Entropy Alloys: A Framework for Model, Interpretation and Multi-Dimensional Visualization"
- Authors: G Song, D Fu, W Chang, Z Fu, L Ma, D Zhang
- Journal: Corrosion Science
- Year: 2025
"A Message Passing Neural Network Framework with Learnable PageRank for Author Impact Assessment"
- Authors: S Guangxuan, FU Dongmei, WU Xiaomeng
- Journal: Advances in Electrical & Computer Engineering
- Year: 2025
"Cross-category prediction of corrosion inhibitor performance based on molecular graph structures via a three-level message passing neural network model"
- Authors: J Dai, D Fu, G Song, L Ma, X Guo, A Mol, I Cole, D Zhang
- Journal: Corrosion Science
- Year: 2022
"From Knowledge Graph Development to Serving Industrial Knowledge Automation: A Review"
- Authors: G Song, D Fu, D Zhang
- Journal: Chinese Control Conference
- Year: 2022
"A Named Entity Extraction Method for Commonly Used Steel Knowledge Graph"
- Authors: Z Ma, L Ma, D Fu, G Song, D Zhang
- Journal: Chinese Intelligent Systems Conference
- Year: 2022