This code repository is for the published paper: An artificial intelligence based data-driven approach for design ideation

Co-authers: Liuqing Chen, Pan Wang, Hao Dong, FengShi, Ji Han, YikeGuo, Peter R.N. Childs, Jun Xiao, ChaoWu

Ideation is a source of innovation and creativity, and is commonly used in early stages of engineering design processes. This paper proposes an integrated approach for enhancing design ideation by applying artificial intelligence and data mining techniques. This approach consists of two models, a semantic ideation network and a visual concepts combination model, which provide inspiration semantically and visually based on computational creativity theory. The semantic ideation network aims to provoke new ideas by mining potential knowledge connections across multiple knowledge domains, and this was achieved by applying “step-forward” and “path-track” algorithms which assist in exploring forward given a concept and in tracking back the paths going from a departure concept through a destination concept. In the visual concepts combination model, a generative adversarial networks model is proposed for generating images which synthesize two distinct concepts. An implementation of these two models was developed and tested in a design case study, which indicated that the proposed approach is able to not only generate a variety of cross-domain concept associations but also advance the ideation process quickly and easily in terms of quantity and novelty.

The architecture of proposed DesignGan:

Details of generator and discirminator: