QUYNH D. TRAN

CASE WESTERN RESERVE UNIVERSITY

Materials Science and Engineering, School of Engineering

Postdoctoral Scholar

About

I received my PhD in Chemistry synthesizing molecules and characterizing them. As a postdoc at SDLE, I am currently working on projects in Additive Manufacturing (AM), Ontology, and Chemistry analyzing data and building AI/ML tools. I’m interested in building AI/ML tools to integrate experiments and research from multiple disciplines. When I am not working, I like to do martial arts (Aikido specifically) and lift weights and hangout with friends.

Research Disciplines

Materials Science

Featured Presentation

See the full schedule here.

Title: Ontology & Historical Data: A Chemical Reaction Case Study for Graph Learning and Semantic Reasoning

Keywords: Ontology, Knowledge Graph, Semantic Reasoning, Chemistry

Abstract: Across industries, more data is generated than ever and presents us with both an opportunity and a challenge. Some challenges include: 1) data from various formats, types, and sources must be integrated, 2) data sharing needs to be seamless as science becomes more interdisciplinary, 3) useful insights, information, and knowledge must be extracted from the vast amount of data to advance science, 4) insights and knowledge must also be accessible to interested non-expert users. At SDLE, the Semantic Web Technology stack (OWL 2 Web Ontology Language (OWL 2), Resource Description Framework (RDF-star), Linked Data (JSON-LD)) is used to address these challenges. We construct mds-Onto (a low-level ontology) to describe and model Materials Science and Chemistry. In essence, mds-Onto enables integration of data from current and past synthesis experiments (more Chemistry) to lifetime and aging experiments (more Materials Science) and researchers to extract useful insights and knowledge. We are working with open-source historical chemical reaction data from Open Reaction Database to FAIRify it into useful RDF graphs stored as JSON-LD. These graphs are AI-ready data for graph deep learning and semantic reasoning to answer many plaguing problems in Chemistry and Materials Science. This work provides a coherent framework to express, share, and represent Materials Science and Chemistry knowledge as knowledge graphs enabling semantic reasoning and graph deep learning. This is applicable to any researcher who is interested in building a semantic data management system to utilize historical and present data more effectively.