Databases Reference
In-Depth Information
There are several surveys related to the Web structure construction on top of
heterogeneous information sources. Kramer
27)
explains technologies to federate
the Web and databases, and reports several case studies. Florescu and others
20)
discuss how database concepts can apply to the Web context. It includes modeling
and querying the Web, information extraction and integration, and Web site
construction and restructuring. Ceri and others
13)
identify ten principles that should
be considered when a Web site managing large amounts of data. Fraternali
22)
surveys
tools and approaches to develop data-intensive Web applications. This work
includes detailed comparison of well-known systems which construct Web
structures based on information sources.
As far as data integration is concerned, a lot of research projects on mediation-
based systems have been conducted so far
17,
24,
29,
31,
40,
42,
45)
. Domenig and Dittrich
16)
classify mediation-based systems and discuss features of the systems in each class.
4
InfoWeaver/AQUA
This section explains our information integration system named
InfoWeaver
(Figure
7) and its visual facility for the Web structure construction. InfoWeaver follows
the mediation approach, and currently accommodates document repositories, the
Web, and relational databases as the underlying information sources. It can
construct Web structures including HTML, XML, and SMIL
48)
documents on
top of the information sources. Its common data model is a hybrid data model
which incorporates ADTs for structured documents in nested relational structures.
A distinguishing feature of the data model is that it has special operators named
converters
to realize
dynamic data conversion
. This mechanism provides a flexible
operational framework for heterogeneous information integration. For example,
relational data can be converted to XML documents and vice versa. Another
feature is its visual Web structure construction facility, named
AQUA
(Amalgamation of QUerying and Authoring)
. It derives both the data integration
and data layout specifications by analyzing user's drag-and-drop-based interactive
visual operations.
Figure 7: InfoWeaver/AQUA
