2022, Jentner, Wolfgang, Sperrle, Fabian, Seebacher, Daniel, Kraus, Matthias, Sevastjanova, Rita, Fischer, Maximilian T., Schlegel, Udo, Streeb, Dirk, Miller, Matthias, Spinner, Thilo, Cakmak, Eren, Sharinghousen, Matthew, Meschenmoser, Philipp, Görtler, Jochen, Deussen, Oliver, Stoffel, Florian, Kabitz, Hans-Joachim, Keim, Daniel A., El-Assady, Mennatallah, Buchmüller, Juri F.

Die COVID-19-Pandemie und ihre rasante Entwicklung innerhalb weniger Wochen stellen völlig neue Anforderungen an die Auswertung von Infektionsstatistiken. CoronaVis stellt interaktive Visualisierungen zur Verfügung, durch die Fallinzidenzen und die Bettenkapazitäten von Intensivstationen (ICUs) in ganz Deutschland analysiert werden können. CoronaVis ist in erster Linie dazu bestimmt, Ärzte, Krisenstäbe und medizinische Entscheidungsträger zu unterstützen und informierte Entscheidungen, zum Beispiel zur Patientenverteilung bei drohender Überlast, zu ermöglichen. CoronaVis skaliert durch flexible Aggregationsmöglichkeiten von der lokalen bis auf die nationale Ebene. Dieser Beitrag stellt die Analysemöglichkeiten von CoronaVis vor und geht näher auf die Leistungsfähigkeit interaktiver Visualisierungen in Hinsicht auf die Unterstützung bei dynamischen Lagen ein.

2018, Cakmak, Eren, Castiglia, Giuliano, Jentner, Wolfgang, Buchmüller, Juri F., Keim, Daniel A.

Control centers for safety-critical infrastructures such as train systems rely on proven, time-tested visualizations to support the decision-making process of the operators. These systems are facing new challenges nowadays as traffic has increased. We describe an incremental visualization design process to adapt Train Management Systems to new tasks, while carefully building on existing techniques to ensure a continuous work environment for operators without or little additional training. The main focus of this work-in-progress is to provide additional contextual information to operators unobtrusively and to incorporate multi-perspective prediction models helping operators to make informed decisions efficiently.

2018, Cakmak, Eren, Schlegel, Udo, Miller, Matthias, Buchmüller, Juri F., Jentner, Wolfgang, Keim, Daniel A.

The VAST Challenge 2018 aims to clarify the situation of the Rose- Crested Blue Pipit’s population in the Boonsong Lekagul Wildlife Preserve. We propose an interactive spectrogram and a representative novel audio glyph to support the classification of bird calls. The second visualization of our system helps to identify spatio-temporal patterns of bird species in the preserve. Using the system, we are able to solve the Mini Challenge 1 (MC1) tasks to classify claimed pipit calls and detect multiple spatio-temporal migratory patterns in the preserve. We find evidence that the pipit population is declining, but we can not identify any clear evidence that the accused company is responsible for the decrease of the pipit population.

2020, Schlegel, Udo, Jentner, Wolfgang, Buchmüller, Juri F., Cakmak, Eren, Castiglia, Giuliano, Canepa, Renzo, Petralli, Simone, Oneto, Luca, Keim, Daniel A., Anguita, Davide

Train operators are responsible for maintaining and following the schedule of large-scale railway transport systems. Disruptions to this schedule imply conflicts that occur when two trains are bound to use the same railway segment. It is upon the train operator to decide which train must go first to resolve the conflict. As the railway transport system is a large and complex network, the decision may have a high impact on the future schedule, further train delay, costs, and other performance indicators. Due to this complexity and the enormous amount of underlying data, machine learning models have proven to be useful. However, the automated models are not accessible to the train operators which results in a low trust in following their predictions. We propose a Visual Analytics solution for a decision support system to support the train operators in making an informed decision while providing access to the complex machine learning models. Different integrated, interactive views allow the train operator to explore the various impacts that a decision may have. Additionally, the user can compare various data-driven models which are structured by an experience-based model. We demonstrate a decision-making process in a use case highlighting how the different views are made use of by the train operator.

2018, Piljek, Isabel, Dehn, Giuliana, Frauendorf, Jannik, Salem, Ziad, Niyazbayev, Yerzhan, Buchmüller, Juri F., Cakmak, Eren, Jentner, Wolfgang, Stoffel, Florian, Keim, Daniel A.

This paper presents our solution to the Mini Challenge 2 (MC2) of the VAST Challenge 2018. We will analyze the provided data set and introduce our visualization tool, which was implemented and tailored to the tasks given by MC2. The tool combines the power of stream graphs, innovative glyph visualizations, box plots, sparklines, heat maps and cross-filter strategies. It allows identifying patterns and anomalies within the provided data set.

2017, Buchmüller, Juri F., Jentner, Wolfgang, Streeb, Dirk, Keim, Daniel A.

In this paper, we present our solution to the VAST Challenge 2017 Mini Challenge 1. We discuss challenges posed by data set and tasks and introduce ODIX, a custom rapid hypotheses testing system tailored to origin-destination data as provided by the challenge. We show findings made with ODIX and illustrate how we apply sequential pattern mining to explore common traffic patterns.

2019, Jentner, Wolfgang, Buchmüller, Juri F., Sperrle, Fabian, Sevastjanova, Rita, Spinner, Thilo, Schlegel, Udo, Streeb, Dirk, Schäfer, Hanna

We present N.E.A.T. - a Visual Analytics approach to the collaborative management of large-scale emergencies. N.E.A.T. unifies the analysis and annotation of heterogeneous, uncertainty-afflicted data sources in a single, adjustable screen. Stakeholders can create individual or shared workspaces providing configurable views tailored to the needs of different emergency responders. Within each workspace, annotated findings are automatically shared in real-time for effective collaboration. We illustrate the functionality of the tool and showcase exemplary findings on the St. Himark incident.

2018, Bäumle, Benedikt, Boesecke, Ina, Buchmüller, Raphael, Metz, Yannick, Buchmüller, Juri F., Cakmak, Eren, Jentner, Wolfgang, Keim, Daniel A.

To solve VAST Mini Challenge 1, we build an interactive visualization tool that allows hypothesis testing and exploratory analysis of the data. The tool contains different visualizations for metadata and audio data analysis. To analyze the recorded bird calls, we trained a Gradient Boosting-classifier to distinguish different bird species. Our tool integrates these results and visualizes them in combination with additional data allowing the user to get context information and confirm the results.

2017, Streeb, Dirk, Buchmüller, Juri F., Schlegel, Udo, Jentner, Wolfgang, Behrisch, Michael, Schneider, Bruno, Seebacher, Daniel

To help ornithologist Mitch in understanding the poor development of the Rose-crested Blue Pipit in terms of the VAST Challenge 2017 Grand Challenge, we apply a diverse set of custom specialized tools and out-of-the-box data analysis systems to a rich data set consisting of satellite images, gas sensor measurements, movement traces and newsletter issues. Following the Visual Analytics approach, we implement a collaborative analysis loop and are able to combine data and gain insights into the current situation of the Boonsong Lekagul Nature Preserve. Finally, we come up with a hypothesis that combines suspect observations to a coherent story of illegal disposal of toxic waste involving two companies located in the reserve's vicinity.