Cognitive visualization for the design of complex systems

• Published in: 2013 IEEE 4th International Conference on Cognitive Infocommunications (CogInfoCom) pp. 363 - 368
• Main Authors: Toth, Janos, Papp, Ildiko, Tornai, Robert, Labancz, Imre, Pocsai, Eniko Hajdune, Hajdu, Andras
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2013
• Subjects: 3D Internet; Accuracy; Algorithm design and analysis; Application generation; Collaboration arena; Customizable content management; Detectors; Information visualization; Servers; Three-dimensional displays; Visualization
• ISBN: 9781479915439; 1479915432
• DOI: 10.1109/CogInfoCom.2013.6719272

In this paper, we introduce a set of new visual representations for complex systems. These visual representations allow for the efficient description of systems that encapsulate a variety of components that can be merged together to form ensembles. These ensemble methods can outperform individual approaches. The whole operation of such systems can be better interpreted using visual tools. Thus, we introduce 3D graphical solutions to visualize the components (e.g. databases, energy functions, algorithms, and voting models) of the system together with their attributes, and also their possible relationships. For a case study, we consider a clinical screening system. As a sufficiently flexible and rich 3D visualization framework, we select the VirCA virtual collaboration arena. The created virtual space supports the manual design of applications from the components with testing them on different databases for the selected energy functions. Moreover, the user can get familiar with the applications generated automatically by machine learning and stochastic search algorithms.