Process-aware Artistic Creativity Evaluation Model and Iterative Generation Algorithm

Abstract

With the rapid advancement of artificial intelligence technology, generative artificial intelligence (AIGC) has seen increasingly wide
spread applications in artistic creation, ranging from image generation to music composition, demonstrating remarkable creative capabilities.
However, existing art evaluation systems predominantly focus on static aesthetic assessments of final outputs, neglecting dynamic evolution dur
ing the creative process, intent expression, and underlying human-computer interaction logic. This often results in evaluation outcomes lacking
profound understanding of the essence of "creativity, "hindering high-quality iterative optimization. This paper proposes a novel "process-aware"
art creativity evaluation paradigm designed to establish dynamic assessment models capable of capturing creative trajectories, thought processes,
and emotional fluctuations, thereby enabling efficient iterative generation algorithms. The study first analyzes limitations of traditional evaluation
methods, systematically expounds the theoretical framework of process awarenessincluding multimodal process data fusion, creative evolution
trajectory modeling, and affective computing mechanismsthen explores an iterative generation algorithm architecture based on this model,with
emphasis on feedback-driven parameter tuning, multi-objective optimization strategies, and human-machine co-evolution mechanisms. Finally,
through concrete application scenarios, the potential and challenges of this model in enhancing intelligent artistic creation and deepening human
machine symbiosis are discussed. This research provides theoretical foundations and technical pathways for overcoming the current "black box"
state of AIGC art creation and achieving paradigm shifts from "output generation"to "process comprehension."