Databases Reference
In-Depth Information
Data
Representation
Presentation
Figure 3.3:
The visualization flow.
Distributed and Inconsistent Data
In a number of Information Visualization (Infovis) applications, it is necessary to integrate and
query data coming from different heterogeneous data sources. This is inherently difficult (and not
especially well researched), and is clearly a challenging activity. Moreover, this data comes in a variety
of types and with different structures: including numeric and non-numeric data, images, videos, and
models. To analyze all this in an integrated fashion is again a challenging activity.
The integration of heterogeneous data is a core activity in database research, and its importance
and topicality are still increasing. However, it is mandatory to understand its role in the context of
Infovis applications. Logic based systems (e.g.,
Calvanese et al.
,
2009
), balancing expressive power
and computational cost, represent the state-of-the-art solutions in data integration, and Infovis can
greatly benefit from such an approach.
The main idea is to evaluate queries through a logic based engine that exploits a conceptual
view of the application domain (i.e., an ontology), rather than a flat description of the data sources.
Such a description, called global schema, is independent of the sources that are mapped through
a logic language into concepts of the global schema. A robust and promising solution is to use
description logics that are a fragment of the first order logic and balance expressive power and
computational cost.
