Can AI predict the future of quantum science?
PODCAST image: Vecteezy
In a recent Physics World podcast, PhD candidate Felix Frohnert from Leiden University explains how his team uncovered hidden connections in scientific literature by using machine learning routines. With this information, it’s possible to predict promising new research ideas in quantum physics. Listen to the podcast now.
Following the evolution of quantum concepts
Researchers from Leiden University use dynamic word embeddings — changing word meanings —that show how quantum ideas are connected across thousands of articles over time. With this, a machine learning model is built that can predict how the field will develop. The model can also suggest which ideas that seem unrelated now may be connected in future research.
Listen to the podcast on PhysicsWorld: Artificial intelligence predicts future directions in quantum science.
A spark for teamwork
Besides predicting new ideas, the analysis also shows possible connections between different parts of the field. This helps researchers find each other and work together, speeding up new discoveries. In this way, AI does not only predict the future of quantum research – it also helps create it.
This research was carried out by Felix Frohnert and Evert van Nieuwenburg, in collaboration with Xuemei Gu and Mario Krenn of the Max Planck Institute.
Read the original article in ‘Machine Learning: Science and Technology’: Dynamic word embeddings for uncovering emerging scientific concepts in quantum physics.