EuroVA19
Permanent URI for this collection
Browse
Browsing EuroVA19 by Author "Andrienko, Natalia"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Contextualized Analysis of Movement Events(The Eurographics Association, 2019) Chen, Siming; Andrienko, Gennady; Andrienko, Natalia; Doulkeridis, Christos; Koumparos, Athanasios; Landesberger, Tatiana von and Turkay, CagatayFor understanding the circumstances, causes, and consequences of events that may happen during movement (e.g., harsh brake, sharp turn), it is necessary to analyze event context. The context includes dynamic attributes of the moving objects before and after the event and external context elements such as other moving objects, weather, terrain, etc. To explore events in context, we propose an analytical workflow including event contextualization, context pattern detection, and exploration of the spatio-temporal distribution of the detected patterns. The approach involves clustering of events based on the similarity of their contexts and interactive visual techniques for exploration of the distribution of the clusters in time, geographic space, and multidimensional attribute space. In close collaboration with domain experts, we apply our method to real-world vehicle trajectories with the purpose of identifying and investigating potentially dangerous driving behaviors.Item Moving Together: Towards a Formalization of Collective Movement(The Eurographics Association, 2019) Buchmüller, Juri; Cakmak, Eren; Andrienko, Natalia; Andrienko, Gennady; Jolles, Jolle W.; Keim, Daniel A.; Landesberger, Tatiana von and Turkay, CagatayWhile conventional applications for spatiotemporal datasets mostly focus on the relation between movers and environment, research questions in the analysis of collective movement typically focus more on relationships and dynamics between the moving entities themselves. Instead of concentrating on origin, destination and the way in between, this inter-mover perspective on spatiotemporal data allows to explain how moving groups are coordinating. Yet, only few visualization and Visual Analytics approaches focus on the relationships between movers. To illuminate this research gap, we propose initial steps towards a comprehensive formalization of coordination in collective movement based on temporal autocorrelation of distance matrices derived from basic movement characteristics. We exemplify how patterns can be encoded using autocorrelation cubes and outline the next steps towards an exhaustive formalization of coordination patterns.