Conceptual Space

Conceptual spaces represent concepts geometrically, aiming to model how humans understand and relate ideas. Current research focuses on learning these spaces from data using various techniques, including neural networks (like variational autoencoders and convolutional neural networks), large language models, and even quantum computing approaches. This work seeks to improve the interpretability and usability of these models for applications such as design creativity assessment, semantic communication, and building more explainable AI systems. Ultimately, advancements in conceptual space modeling promise to enhance our understanding of human cognition and lead to more robust and human-centered AI.