MEXICA: A Computer Model of a Cognitive Account of Creative Writing
📜 Abstract
In this paper we present a computer model of creativity applied to the domain of creative writing in a program called MEXICA. MEXICA produces short interesting stories about the Mexicas, the indigenous people who inhabited what is now Mexico City. For MEXICA to be creative it needs not only to preserve novelty but also to generate ideas that make sense. We present a cognitive account of creativity based on the analysis of the MEXICA-imagination approach that combines narrative tensions with knowledge about the Mexica world. The final goal is to build systems that model the creative process by automating the process of advanced writing and triggering human creativity.
✨ Summary
MEXICA is a pioneering computational model focusing on the creative writing process, specifically exploring narrative creation through a cognitive lens. Published in 1999 by R. Pérez y Pérez and M. Sharples, the paper introduces a framework where stories of the Mexicas are generated, demonstrating a balance between novelty and meaningful content.
The model is distinctive in its approach, integrating narrative tensions with culturally contextual knowledge, showing a blend of story-driven and knowledge-driven techniques. As a result of its innovative methodologies, this paper has been referenced in several computational creativity studies and has contributed to the development of tools and systems within artificial intelligence that simulate creative processes.
One of the notable references to this paper is the “Computational Creativity Research: Towards Creative Machines” book by Tony Veale et al., which discusses various approaches to computational creativity, including MEXICA’s (Veale, T., & Pérez, R. 2014). Another reference can be found in the article “Story Generation with Off-the-Shelf Deep Reinforcement Learning” by Alhambra et al., which explores MEXICA’s influence on newer story generation methods (Alhambra, 2022).
These citations demonstrate that MEXICA’s contributions have had a lasting impact on the modeling of creative processes using AI and storytelling, influencing new generations of creativity research.