Extraction and visually driven analysis of geography and dynamics of people's reaction to events (EVA-VGI)

  • Dr. Eva Hauthal
    Dr. Eva Hauthal
    Dresden
  • Ross Purves
    Ross Purves
    Zürich
  • Prof. Dr. Stefan Wrobel
    Prof. Dr. Stefan Wrobel
    Bonn
  • Dirk Burghardt
    Prof. Dr.-Ing. Dirk Burghardt
    Dresden
  • Alexander Dunkel
    Dr.-Ing. Alexander Dunkel
    Dresden
  • Prof. Dr. Gennady Andrienko
    Prof. Dr. Gennady Andrienko
    Bonn
  • Prof. Dr. Natalia Andrienko
    Prof. Dr. Natalia Andrienko
    Bonn
EVA-VGI

The rise of so-called Volunteered Geographic Information (VGI) has brought with it fundamental changes not only in the nature of geographic data, but also its production and accessibility. These changes have implications across the board for those carrying out research requiring geographic data, and most profoundly for research exploring how humans interact with and are affected by changes in their environment. In this project we propose moving beyond event detection, to investigate ways in which events can be characterised according to four key dimensions: spatial, temporal, thematic and social context. This characterisation will allow us to capture information about how groups of users are both affected by, and react to events and their evolution. To do so will require us to relate a variety of data sources with heterogeneous properties, which in some cases are generated with high velocities and volumes and have multiple granularities.

Our approach to characterising events has the following central pillars: data extraction and event identification; analysis of data content and semantics and associated interactive, dynamic visualisation. Analysis methods typical generate more information, and as such, given the large volumes of data that are our starting point will require effective visual analytics approaches enabling users to interact with the expected large data volumes. Our objective is therefore to develop a system of integrated visual and computational methods enabling investigation of people’s reactions to significant events through analysis of volunteered geographic information containing both explicit and implicit georeferences. Thus we will firstly develop a taxonomy of events focusing on the four mentioned dimensions associated with events and ways in which events are reacted to. The extraction of events from VGI will be based on a combination of spatial, temporal and textual analysis methods to link events, groups and their characteristics to space, time and social context. Our analysis will focus on reactions to events, and in so doing the socio-demographics of involved groups. In the analysis of the social aspects of group reactions to events a central tenet will be the use of privacy preserving methods which effectively obfuscate individuals.

We will illustrate our concepts through the implementation of a set of demonstrators including (but not limited to) disaster management, public involvement in local affairs, migration, climate and environment, social geography and medicine and health. To achieve our aims we have brought together a project team with expertise in investigating VGI, for example exploring the spatio-temporal behaviour from such data, describing the semantics of specific and generic aspects of place, and analysing affective data associated with location.

Publications

  1. Hauthal, E., Burghardt, D., & Dunkel, A. (2019). Analyzing and Visualizing Emotional Reactions Expressed by Emojis in Location-Based Social Media. International Journal of Geo-Information, 8(3), 21. DOI: 10.3390/ijgi8030113
  2. Dunkel, A., Andrienko, G., Andrienko, N., Burghardt, D., Hauthal, E., & Purves, R. (2018). A conceptual framework for studying collective reactions to events in location-based social media. International Journal of Geographical Information Science, 1–25. DOI: 10.1080/13658816.2018.1546390
  3. Löchner, M., Dunkel, A., & Burghardt, D. (2019). Protecting privacy using HyperLogLog to process data from Location Based Social Networks. LESSON.
  4. Löchner, M., Dunkel, A., & Burghardt, D. (2018). A privacy-aware model to process data from location-based social media. VGI Geovisual Analytics Workshop.