Global Positioning System Reference
In-Depth Information
enough computer science specialists willing to accept the challenge
and learn about spatial data, its manipulation, and problems.
• Most of the projects related to spatial data require the knowledge of
geo-specialists and they are usually more familiar with GIS products
with their traditional data storage than spatial DBMSs.
• The common supposition that spatial DBMSs are “heavy”, i.e., map
retrieval is slower, prevails among geo-specialists and to our knowledge
there are no publications that contradict this assumption.
• Web map services that do not require DBMSs are currently very
popular.
• There is no evaluation (i.e., benchmarking) for spatial DBMSs or a
guide for choosing one according to specifi c needs, e.g., economic
factor, performance, storage, ease of use, compliance with standards,
among others.
Spatial ETL
The practical importance of the ETL processes is high, since this time-
consuming process is responsible for data integration and customization.
Several works refer to conceptual modeling of ETL processes for conventional
data (Akkaoui et al. 2011; Albrecht and Nauman 2008; Silva et al. 2012;
Trujillo and Luján-Mora 2003; Vassiliadis et al. 2005; 2009). These proposals,
sometimes complex in nature, intend to provide formal foundations and
vendor-independent models for the design of ETL processes. To the best
of our knowledge, there is not a formal proposal for spatial ETL processes.
Some works refer to specifi c problems, e.g., integrating spatial data related
to water quality (Wang et al. 2010) or provide data for health resource and
service administration (Cohen and Baitty 2005). As a consequence, spatial
ETL processes are usually implemented using specifi c tools according to
the problem at hand.
ETL processes not only integrate data and transform it according to
user analysis requirements, but they also are important in improving spatial
data quality. If data collectors are not aware of data quality and systems
implementers are not familiar with the methods to improve quality, the
last element of the chain, i.e., data customers, obtain incorrect or defi cient
data that can be harmful for the company/organization, whether this data
is operational or decisional (Talhofer et al. 2011). However, traditionally
the digitalization and analysis of spatial data relied on geo-specialists who
were aware that GISs would not help them improve spatial data quality
since these systems work under the assumption that data is perfect and do
not provide the capabilities to establish data quality control (Delavar and
Devillers 2010). In contrast, currently, different users with a profi le that is
usually unrelated to GISs are in charge of either collecting data from existing
Search WWH ::
Custom Search