Abstract
Computational creativity researchers have long been searching for a reliable creative method of generating transformational creativity in Creativity Support Tools, in vain, especially when these systems are supposed to take in a user's unfinished creative work and produce representational and creative outputs as continuations to the user input. In this paper we propose a new creative method called Conceptual Recombination to take up this challenge. We first define creative work for this study followed by creative work ontology to be the theoretical background of Conceptual Recombination. We further refer to application ontology and regard Conceptual Recombination as the task model for creative work ontology. In this task model there are three levels of prediction leading to the formations of output features, output structures, and their combinations as the final system outputs constrained by rules, biases, and homeomorphism. Furthermore, this new creative method allows the use of exploratory creativity on structures and transformational creativity on features to attain a balance between usefulness and novelty in system outputs. A 7-tuple computational model and the search mechanisms for exploratory and transformational creativity are also defined for it. Lastly, we evaluate Conceptual Recombination with our case study about producing a 2-dimensional asymmetrical shape with a given symmetrical shape to demonstrate its practicality and conclude that it not only offers a new reliable creative method for Creativity Support Tools, but also provides an objective evaluation method for transformational creativity.
Original language | English |
---|---|
Pages (from-to) | 1-12 |
Number of pages | 12 |
Journal | Knowledge-Based Systems |
Volume | 53 |
DOIs | |
Publication status | Published - 1 Nov 2013 |
Keywords
- Computational creativity
- Conceptual Recombination
- Creative method
- Creative work ontology
- Creativity Support Tools
ASJC Scopus subject areas
- Software
- Management Information Systems
- Information Systems and Management
- Artificial Intelligence