Integrating Crowdsourcing & Gamification in an Automatic Architectural Synthesis Process

Abstract

This work covers the methodological approach that is used to gather information from the wisdom of crowd, to be utilized in a machine learning process for the automatic generation of minimal apartment units. The flexibility in the synthesis process enables the generation of apartment units that seem to be random and some are unsuitable for dwelling. Thus, the synthesis process is required to classify units based on their suitability. The classification is deduced from opinions of human participants on previously generated units. As the definition of ``suitability'' may be subjective, this work offers a crowdsourcing method in order to reach a large number of participants, that as a whole would allow to produce an objective classification. Gaming elements have been adopted to make the crowdsourcing process more intuitive and inviting for external participants.