Visual Analytics: Definition, Process, and Challenges - Information Visualization: Human-Centered Issues and Perspectives - page 159

Fig. 3. Visual Analytics integrates Scientific and Information Visualization with core
adjacent disciplines: Data management and analysis, spatio-temporal data, and human
perception and cognition. Successful Visual Analytics research also depends on the
availability of appropriate infrastructure and evaluation facilities.
An example for a common problem in several of the disciplines is that of scalabil-
ity with data size. The larger the data set to be handled gets, the more dicult
it gets to manage, analyze, and visualize these data effectively. Researching ap-
propriate forms to represent large data volumes by smaller volumes containing
the most relevant information benefits each of the data management, analy-
sis, and visualization fields. On top of these individual progresses, a synergetic
collaboration of all these fields may lead to significantly improved processing
results. Consider a very large data stream. Appropriate data management tech-
nology gives ecient access to the stream, which is intelligently processed and
abstracted by an automatic analysis algorithm which has an interface to the
data management layer. On top of the analysis output, an interactive visual-
ization which is optimized for ecient human perception of the relevant infor-
mation allows the analyst to consume the analysis results, and adapt relevant
parameters of the data aggregation an analysis engines as appropriate. The com-
bination of the individual data handling steps into a Visual Analytics pipeline
leads to improved results and makes data domains accessible which are not ef-
fectively accessible by any of the individual data handling disciplines. Similar
argumentations apply to other related fields and disciplines. In many fields, vi-
sualization is already used and developed independently as a means for analyzing
the problems at hand. However, a unified, interdisciplinary perspective on using
visualization for analytical problem-solving will show beneficial for all involved
disciplines. As common principles, best practices, and theories will be developed,
these will become usable in the individual disciplines and application domains,
providing economies of scale, avoiding replication of work or application of only
sub-optimal techniques.