Project Details
Projekt
Visual feature space analysis
Applicant
Professor Dr. Tobias Schreck
Subject Area
Security and Dependability, Operating-, Communication- and Distributed Systems
Term
from 2008 to 2014
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 81159851
Complex data arises in many application domains. Visual analysis in complex data types such as time series or multimedia data is not straightforward but requires definition of appropriate similarity functions, often based on high-dimensional descriptors. The identification of descriptors suited for the visual analysis task at hand is not straightforward. In the first phase of our SPP part project, we research methodologies to embed the descriptor choice problem into selected Visual Analytics applications supporting single (individual) complex data types.Often, the data set under concern not only consists of a single complex data type, but is aggregated (composed) of several complex data types which are mutually interrelated. An example are repositories of research data, where data types such as temporal, spatial, image, or textual data are jointly representing factual information from experimentation or observation. We will research new Visual Analytics approaches for data sets aggregated of multiple complex data types, for each of which in turn multiple descriptors are given. Methodological questions on visual and algorithmic design will be addressed, and applications for selected use cases will be devised and evaluated. We will consider content-based visual search and cluster analysis applications for aggregated complex data, supported by automatic means of guiding the user through the analysis and retrieval space.
DFG Programme
Priority Programmes
