IN CHEMISTRY
, Symbolic AI systems
Symbolic AI, the first form of AI pioneered in the 1960s and
1970s, is characterized by an explicit representation of domain
knowledge as formal facts, rules, and logical structures that
translates human understanding into a computer-
understandable representation that mimics human-expert
reasoning. In reaction modeling, such systems have been
developed by utilizing the formalization of chemistry rules
through reaction transforms that map the changes in atom-
connectivity during reactions.
After forty years of research, in fits and starts, on developing
intelligent systems for reaction prediction and chemical
synthesis, we seem to have finally arrived at an exciting and
productive phase thanks to the recent progress in artificial
intelligence. In this perspective, we have reviewed the recent
surge of artificial intelligence methods in the area of reaction
prediction and chemical synthesis, highlighting the current
challenges and opportunities. Despite the quick and easy
success.