An exploratory method for an alternative narrative of housing history in Istanbul

Abstract

Analysis and visualization of spatial data has proven to be an effective tool in communicating apparent or latent spatial information in an efficient way. Here a specific data processing and visualization method is employed in creating an original base for a nonlinear historical narrative of a specific urban phenomenon. Housing development in Istanbul is selected as a case to be explored through a heuristic method involving correspondence analysis (COA) complemented by clustering methods and Bertin's graphics theory. COA is an exploratory method for cross-tabular data analysis, which facilitates the interpretation of data by visualizing the relative relationship between its variables. Two sets of data on the shares of the state, the private sector, and housing cooperatives in housing development in Istanbul between 1987 and 2007 have been processed by COA. The outputs of the process are correspondence maps and Bertin graphics. Maps create a basis for a spatiotemporal analyses of public and private housing development in Istanbul, while revealing certain relation networks and breaks in housing history that are not quite perceptible by looking at the data tables only. Thus, drawing upon the discoveries from the maps, a nonlinear narrative of housing history is proposed as a collection of several thorough analyses of the discoveries within the economic, political, and social setting particular to this geography.