Towards innovative interfaces for historical and geographical thinking: qualitative filtering with co-occurrence graphs for geohumanities

摘要

The prototype described in this paper used dynamic qualitative filtering on multiple linked views to explore the social-spatial collective memory of places as described by co-occurrence networks that emerge from Instagram posts. We achieved this by creating a geographic extension to an existing humanities tool for network analysis of co-occurrence in large-scale multimedia datasets. This enabled the networks to be explored on multiple geographic (near and far) and textual (close and distant) scales to understand the networks of people, themes and places that are mentioned in the same posts. The aim was to understand if such a prototype could help disentangle the social construction of place as represented by peoples?social media traces that are shared on Instagram and analysed by combing historical and geographical thinking. To test the validity of the prototype, we took one of the many geographical theories of place and investigated how it can be applied and evaluated using the tool with an Instagram dataset. Taking as our basis Cresswell theory of places, which supposes that places are an embodiment of materiality, meaning and practice, we started unravelling the complexity of understanding the sense of place with respect to the Via Francigena cultural route. Using the geographically modified version of the tool we investigate the limitations of using the technology, text analysis processes and a technology mediated dataset for such purposes. The result is a much deeper understanding of the interface needs and data processing and analysis requirements for tools that foster historical and geographical thinking