10.1184/R1/7182263.v1 Yingxiu Lu Yingxiu Lu Conversational Form-Generation: An Application of Interactive Genetic Algorithm to Architectural Design Carnegie Mellon University 2018 Interactive Genetic Algorithm(IGA) Conversational Design Computational Design 2018-05-15 00:00:00 Thesis https://kilthub.cmu.edu/articles/thesis/Conversational_Form-Generation_An_Application_of_Interactive_Genetic_Algorithm_to_Architectural_Design/7182263 The design process can be regarded as a recursive conversation between representation and interaction, where designers recursively "generate representations of their<br>interactions, and by interacting with representations simultaneously, designers generate relations with the representations of which they can then interact and repeat this process recursively." (Maturana and Varela, 1991)<br>This recursive conversation is quite typical and essential in the form-generation stage. This thesis shows how this recursive conversation can be boosted using Interactive<br>Genetic Algorithm(IGA). With the assistance of IGA, the form-generation software can support an interactive conversation between designers and computational<br>tools, by taking designers’ evaluation and modification into consideration during computation process. Based on this interactivity of IGA, this thesis proposes an alternative<br>working relationship between designers and computational tools, where they work in different scopes parallelly and share the outcome regularly. This alternative<br>relationship offers an opportunity where designers could maximize their creativity, and computational tools could apply their computation capability to optimize the<br>design idea. A form-generation platform is implemented grounded on the proposed working relationship, for form study in the early stage of design process.<br>