Visual Search in Research Data

Due to technical advances in acquisition, processing and storage of primary research data, in domains such as Meteorology, Earth Observation, and Climate Simulation, increasing amounts of primary research data are collected. Data sets being of high value for current and future research are already often stored and made available by data center organizations, and are indexed by scientific libraries. While these trends improve the transparency and availability of scientific research data, access to these data sets is to date mainly based on metadata information. The goal of this project is to research content-based visual search and presentation techniques for user-friendly accessing and exploration of large collections of time-oriented research data sets. The aim is together with established metadata-based access methods to improve the overall accessibility of the indexed data sets. To this end, concepts such as a visual cataloguing and query-by-example and -sketch will be developed for collections of scientific time series. A specific challenge is seen in the appropriate combination of content- and metadata-based searching modalities.

Contact: Tatiana von Landesberger,Maximilian Scherer
Grant by: the Leibniz Gemeinschaft under the SAW program.
Collaboration partners: German National Library of Science and Technology (TIB) and Fraunhofer Institute for Computer Graphics Research.
Data repository: PANGAEA® - Data Publisher for Earth & Environmental Science.
Data type: Time-oriented research data, provided by the Baseline Surface Radiation Network (BSRN).


A visual catalog of a scientific time series dataset. Using a visual cluster algorithm, thousands of daily temperature curves from different stations all over the world can be arranged in one visualization.

Time series sketch editor. the user can select example patterns or create individual sketches of time series patterns to formulate a content-based query. The application executes the user-defined query and provides a result set of the most similar time series patterns from our index structure.

Visual catalog with a search result visualization. the colormap indicates similar time series patterns with blue colors, unsimilar patterns are denoted with red color values. The color gradient is highly homogeneous because the used cluster algorithm preserves the topologic ordering of the time series patterns.

Filtering the visual catalog. Visualization of occurrence of certain keywords from associated metadata. The white-yellow colormap indicates the number of patterns of each cluster cell that correspond to the filter query (also called density histogram). Yellow colors denote cluster cells with high density.

Publications

TimeSeriesPaths: Projection-Based Explorative Analysis of Multivarate Time Series Data.
J. Bernard, J Wilhelm, M. Scherer, T. May and T. Schreck
In: Proc. Int. Conference in Central Europe on Computer Graphics, Visualization and Computer Vision,2012
A Benchmark for Content-Based Retrieval in Bivariate Data Collections.
M. Scherer, T. von Landesberger and T. Schreck
In: Proc. Int. Conference on Theory and Practice of Digital Libraries, to appear, 2012
Guided Discovery of Interesting Relationships Between Time-Series Clusters and Metadata Properties.
J. Bernard, T. Ruppert, M. Scherer, T. Schreck and J. Kohlhammer
In: Proc. Int. Conference on Knowledge Management and Knowledge Technologies, Special Track on Theory and Applications of Visual Analytics, ACM ICPS, to appear, 2012.
Content-Based Layouts for Exploratory Metadata Search in Scientific Research Data.
J. Bernard, T. Ruppert, M. Scherer, J. Kohlhammer and T. Schreck
In: Proc. ACM/IEEE Joint Conference on Digital Libraries, pages 139-148, 2012.
A Visual Digital Library Approach for Time-Oriented Research Data
J. Bernard, J. Brase, D. Fellner, O. Koepler, J. Kohlhammer, T. Ruppert, T. Schreck, I. Sens
In: Proceedings of Springer International Journal of Digital Libraries, ECDL 2010 Special Issue, 2011.
Multiscale Visual Quality Assessment for Cluster Analysis with Self-Organizing Maps
J. Bernard, T. von Landesberger, S. Bremm, and T. Schreck
In: In IS&T/SPIE Conference on Visualization and Data Analysis, pages 78680N.1 - 78680N.12. SPIE Press, 2011
Micro-Macro Views for Visual Trajectory Cluster Analysis
J. Bernard, T. von Landesberber, S. Bremm, T. Schreck
In: IEEE Information Visualization, 2009
Visual cluster analysis of trajectory data with interactive kohonen maps
T. Schreck, J. Bernard, T. Von Landesberger, J. Kohlhammer In: Information Visualization 8(1) 14 - 29, 2008  

Contact

Technische Universität Darmstadt

Interactive Graphics Systems Group

Fraunhoferstr. 5
64283 Darmstadt

Tel. +49 6151 155 679

icon email office@gris.tu-

A A A | Print Print | Legal note Legal Notes | Contact Contact
to topto top