Database Reference
In-Depth Information
Scalability with Data Volumes and Data Dimensionality:
Visual Ana-
lytics techniques need to be able to scale with the size and dimensionality of
the input data space. Techniques need to accommodate and graphically repre-
sent high-resolution input data as well as continuous input data streams of high
bandwidth. In many applications, data from multiple, heterogeneous sources
need to be integrated and processed jointly. In these cases, the methods need
to be able to scale with a range of different data types, data sources, and levels
of quality. The visual representation algorithms need to be ecient enough for
implementation in interactive systems.
Quality of Data and Graphical Representation:
A central issue in Visual
Analytics is the avoidance of misinterpretations by the analyst. This may result
due to uncertainty and errors in the input data, or limitations of the chosen
analysis algorithm, and may produce misleading analysis results. To face this
problem, the notion of data quality, and the confidence of the analysis algorithm
needs to be appropriately represented in the Visual Analytics solutions. The user
needs to be aware of these data and analysis quality properties at any stage in
the data analysis process.
Visual Representation and Level of Detail:
To accommodate vast streams
of data, appropriate solutions need to intelligently combine visualizations of
selected analysis details on the one hand, and a global overview on the other
hand. The relevant data patterns and relationships need to be visualized on
several levels of detail, and with appropriate levels of data and visual abstraction.
User Interfaces, and Interaction Styles and Metaphors:
Visual Analytics
systems need to be easily used and interacted with by the analyst. The analyst
needs to be able to fully focus on the task at hand, not on overly technical or
complex user interfaces, which potentially distract. To this end, novel interaction
techniques need to be developed which fully support the seamless, intuitive visual
communication with the system. User feedback should be taken as intelligently
as possible, requiring as little manual user input as possible, which guarantees
the full support of the user in navigating and analyzing the data, memorizing
insights and making informed decisions.
Display Devices:
In addition to high-resolution desktop displays, advanced
display devices such as large-scale power walls and small portable personal assis-
tant, graphically-enabled devices need to be supported. Visual Analytics systems
should adapt to the characteristics of the available output devices, supporting
the Visual Analytics workflow on all levels of operation.
Evaluation:
Due to the complex and heterogeneous problem domains addressed
by Visual Analytics, so far it has been dicult to perform encompassing evalua-
tion work. A theoretically founded evaluation framework needs to be developed
which allows assessing the contribution of any Visual Analytics system toward
the level of effectiveness and eciency achieved regarding their requirements.
Search WWH ::
Custom Search