Towards Qualitative Geovisual Analytics: A Case Study Involving Places, People, and Mediated Experience

Abstract

This article seeks to address the gap between the quantitative, summative data that are typically engaged in geovisual analytics and the more personal, experiential ways of knowledge construction accentuated by qualitative GIS. By incorporating diverse forms of data within a high-dimensional conceptual framework, we set out a type of qualitative geovisual analytics. This approach is attentive to the epistemological limitations of singular data sources and highlights the multiple ways of exploring neighbourhoods. The article reports on a project that used an online survey, including collection of personal impressions of San Diego neighbourhoods based on street-level video. Three attribute spaces are conceptualized: survey respondents' characteristics, attributes of San Diego neighbourhoods, and characteristics of the words used to describe these neighbourhoods. The self-organizing map (SOM) technique was used to reduce the dimensionality of these attribute spaces for visual exploration. This approach makes several important contributions, including a demonstration of “scaling up” the work that has been done in qualitative GIS. It productively integrates experiential data with a geovisual analytics approach and reaffirms the multiple meanings of visualization.