Foundational Study on Ontology Based on Bayesian Network for Design Intent & Elements using Simple Invitation Card Design

Abstract

Background Artificial intelligence) (AI) is a branch of machine learning that facilitates the design and enhances designers efficiency. AI has increased the assistants efficiency in searching appropriate images. With the rate AI is developing, new tools will be developed that co-work with a designer to help apprentices learn the designers style and understand the design intent. Therefore, reasoning design intent and design elements are the important foundation for AI-based design tools. The purpose of this study is to provide such a foundation by proposing a model to infer the relationship between design intent and elements. Methods We observed and recorded the design intent and how a designer designs with elements. Then, we analyzed causal relations and probability between design intent and elements according to Bayes' theorem. In addition, we constructed an ontology to generate the Bayesian network. Finally, we simulated design intent reasoning based on the selected designers elements. Result We proposed an ontology and Bayesian Network for the design intent from the selected elements and provided appropriate design elements recommendations for specific design intent. Conclusion We built an ontology to make programs understand the relationship between design intent and design elements based on the observation of a designers work. We proposed a Bayesian Network to reason the designers intent from the designers selected elements and offer design suggestions based on the design intent. This study is a foundational study for the development of AI design tools that co-work with a designer like apprentices.