This paper explores the application of Formal Learning Theory (FLT) to the field of Computational Creativity (CC), proposing a new framework called Formal Creativity Theory (FCT). The authors introduce key concepts from FLT, such as languages, grammars, texts, and scientists, and reinterpret them in the context of creativity.
The paper defines several creativity-related terms within the FCT framework, including artefacts, experience and fate sequences, conceptual spaces, inspiring sets, hypothetical conceptual spaces, platonic conceptual spaces, and rules. It then introduces two specific properties of artefacts - novelty and transformativeness - and analyzes their relationship.
The authors prove that in the general class of scientists, novelty is neither necessary nor sufficient for transformational creativity. However, they show that for a subclass of "set-driven" scientists, novelty is necessary for transformativeness. This suggests that assuming set-driven properties may constitute a significant change in the creativity modeling paradigm, with important implications.
The paper concludes by outlining future research directions, such as exploring alternative property definitions, modeling other creativity concepts, and investigating the computability of these properties. It also emphasizes the importance of revitalizing fundamental theoretical research in Computational Creativity, alongside practical applications.
To Another Language
from source content
arxiv.org
Дополнительные вопросы