Database Reference
In-Depth Information
Infrastructure: Managing large amounts of data for visualization or analysis
requires special data structures and mechanisms, both in memory and disks.
Achieving interactivity means refreshing the display in 100ms at worst whereas
analyzing data with standard techniques such as clustering can take hours to
complete. Achieving the smooth interaction required by the analysts to perform
their tasks while providing high-quality analytical algorithms need the combi-
nation of asynchronous computation with hybrid analytical algorithms that can
trade time with quality. Moreover, to fully support the analytical process, the
history of the analysis should also be recorded and interactively edited and an-
notated. Altogether, these requirements call for a novel software infrastructure,
built upon well understood technologies such as databases, software components
and visualization but augmented with asynchronous processing, history manage-
ments and annotations.
7 Examples for Visual Analytics Applications
7.1 Visual Analytics Tools for Analysis of Movement Data
With widespread availability of low cost GPS devices, it is becoming possible to
record data about the movement of people and objects at a large scale. While
these data hide important knowledge for the optimization of location and mobil-
ity oriented infrastructures and services, by themselves they lack the necessary
semantic embedding which would make fully automatic algorithmic analysis pos-
sible. At the same time, making the semantic link is easy for humans who however
cannot deal well with massive amounts of data. In [5] we argue that by using
the right visual analytics tools for the analysis of massive collections of move-
ment data, it is possible to effectively support human analysts in understanding
movement behaviors and mobility patterns.
Figure 5 shows a subset of raw GPS measurements presented in so-called
space-time cube. The large amount of position records referring to the same
territory over a long time period makes it virtually impossible to do the analysis
by purely visual methods.
The paper [5] proposes a framework where interactive visual interfaces are
synergistically combined with database operations and computational process-
ing. The generic database techniques are used for basic data processing and ex-
traction of relevant objects and features. The computational techniques, which
are specially devised for movement data, aggregate and summarize these objects
and features and thereby enable the visualization of large amounts of informa-
tion. The visualization enables human cognition and reasoning, which, in turn,
direct and control the further analysis by means of the database, computational,
and visual techniques. Interactive visual interfaces embrace all the tools.
Thus, in order to detect and interpret significant places visited by the mov-
ing entities, the positions of stops are extracted from the data by means of
appropriate database queries. Then, clustering methods are applied to detect
frequently visited places. Interactive visual displays put the results in the spa-
tial and temporal contexts. The spatial positions of the stops can be observed on
Search WWH ::




Custom Search