Visual Analytics: Definition, Process, and Challenges - Information Visualization: Human-Centered Issues and Perspectives - page 160

3.1 Visualization

Visualization has emerged as a new research discipline during the last two dec-
ades. It can be broadly classified into Scientific and Information Visualization.
In Scientific Visualization, the data entities to be visualized are typically 3D
geometries or can be understood as scalar, vectorial, or tensorial fields with ex-
plicit references to time and space. A survey of current visualization techniques
can be found in [22,35,23]. Often, 3D scalar fields are visualized by isosurfaces or
semi-transparent point clouds (direct volume rendering) [15]. To this end, meth-
ods based on optical emission- or absorption models are used which visualize the
volume by ray-tracing or projection. Also, in the recent years significant work
focused on the visualization of complex 3-dimensional flow data relevant e.g.,
in aerospace engineering [40]. While current research has focused mainly on e-
ciency of the visualization techniques to enable interactive exploration, more and
more methods to automatically derive relevant visualization parameters come
into focus of research. Also, interaction techniques such as focus&context [28]
gain importance in scientific visualization.
Information Visualization during the last decade has developed methods
for the visualization of abstract data where no explicit spatial references are
given [38,8,24,41]. Typical examples include business data, demographics data,
network graphs and scientific data from e.g., molecular biology. The data con-
sidered often comprises hundreds of dimensions and does not have a natural
mapping to display space, and renders standard visualization techniques such as
( ) plots, line- and bar-charts ineffective. Therefore, novel visualization tech-

niques are being developed by employing e.g., Parallel Coordinates and their
numerous extensions [20], Treemaps [36], and Glyph [17]- and Pixel-based [25]
visual data representations. Data with inherent network structure may be visual-
ized using graph-based approaches. In many Visualization application areas, the
typically huge volumes of data require the appropriate usage of automatic data
analysis techniques such as clustering or classification as preprocessing prior to
visualization. Research in this direction is just emerging.

x, y

3.2 Data Management

An ecient management of data of various types and qualities is a key com-
ponent of Visual Analytics as this technology typically provides the input of
the data which are to be analyzed. Generally, a necessary precondition to per-
form any kind of data analysis is an integrated and consistent data basis [18,19].
Database research has until the last decade focused mainly on aspects of e-
ciency and scalability of exact queries on homogeneous, structured data. With
the advent of the Internet and the easy access it provides to all kinds of hetero-
geneous data sources, the database research focus has shifted toward integration
of heterogeneous data. Finding integrated representation of different data types
such as numeric data, graphs, text, audio and video signals, semi-structured
data, semantic representations and so on is a key problem of modern database