Knowledge Engineering (KE) is a field of artificial intelligence that deals with developing and maintaining intelligent systems capable of solving complex problems using symbolic methods. These systems mimic the decision-making process of a human expert, thereby enhancing the quality and efficiency of problem-solving in various domains.
The research in KE focuses on the chemical domain, where we use ontologies to represent knowledge. Ontologies are digital frameworks that enable the instantiation of heterogeneous chemical information in a machine-readable format. This process involves structuring the information in a manner that facilitates efficient and accurate reasoning.
We establish a repository of valuable chemical information and knowledge by populating the knowledge graph with instantiated knowledge. This knowledge graph serves as the backbone for our problem-solving systems, allowing us to build software agents that replicate the decision-making processes of domain experts. These agents use evidence-based reasoning to access the knowledge graph, query information and generate new knowledge.
Our work results in a system capable of solving complex chemical problems more efficiently and effectively than traditional methods. By mimicking the decision-making processes of a human expert, our software agents enable us to automate complex decision-making tasks, improving the quality of decision-making while reducing the time required to solve complex problems.
Knowledge Engineering in Chemistry: From Expert Systems to Agents of Creation
A. Kondinski, J. Bai, S. Mosbach, J. Akroyd, M. Kraft
Acc. Chem. Res. 2023, 56, 2, 128-139.
Automated Rational Design of Metal-Organic Polyhedra
A. Kondinski, A. Menon, D. Nurkowski, F. Farazi, S. Mosbach, J. Akroyd, M. Kraft
J. Am. Chem. Soc. 2022, 144, 26, 11713–1172.