This week we are sharing Tamar Ofer’s work in “Hacking Housing”, under the Computation Design Lab. @tamaraofer is a pursuing a dual degree in SMArchS Urbanism and MSRED. How can Architects utilize market research for a user-focused design? How can the digital marketplace serve architects as a rapid design tool to learn from their users? Architects need quicker feedback loops, earlier in the design process. This experiment lays the groundwork for a novel floorplan analysis approach leveraging the digital property marketplace for crowdsourcing user preference. It proposes a new design research strategy for measuring preferences and the importance users place on different features of spatial configuration in residential layouts. To do so, crowdsourcing methods were conducted to evaluate spatial preference via a joint geospatial preference and pairwise ranking (i.e. rental listing inquires) of three rental unit type couples (six listings in total). Overall, 7,467 users took part in the three-week experiment, and 376 (5% of total viewers) inquired for further details as potential tenants. Given a vectorized input floor plan via MIT’s Data Warehouse, input images are segmented into different room regions, generating a room graph layout to encode the geometric and semantic information. these are then manipulated via restructuring spatial configuration from the typical clustered organization to a linear circulation of cells (e.g. rooms). After the coupling process is complete, the “twin layouts” have then been put into action via three different marketplace rental platforms as the testing grounds for the target user preference evaluation. This experiment illustrates the feasibility and efficacy of crowdsourcing floorplan comparative analysis for user preference evaluation based on digital rental marketplaces for rapid user-based spatial valuation - replacing rules of thumb with accurate measures to determine how specific users evaluate spatial typology and thus can improve our space by gathering and analyzing member feedback. Instructor: Prof. Takehiko Nagakura, Dr. Daniel Tsai & Prof. Guzden Varinlioglu. TA: Charles Wu. @tamaraofer