Database Reference
In-Depth Information
2.6 RECENT TRENDS IN THE FIELD
Recent trends in Semantic Web technologies have taken several different avenues.
From the “Web 1.0” direction, there has been incremental change toward embedding
structured information into Web pages, using microformats. Microformats reuse
standard XHTML tags to express metadata, which can provide a low barrier to entry
for organizations that are put off by the complexities of description logics and knowl-
edge modeling in heavier-weight Semantic Web technology. For those who are reluc-
tant to invest in a completely different way of managing their business intelligence or
content, the microformat option offers a first, lower-risk step toward greater semantic
understanding of their Web content, based on the “pave-the-cowpaths” principle.
There are several microformats for encoding geographical information, for
example, to mark up WGS84 geographic coordinates (latitude and longitude),
15
or encoding information about waypoints,
16
which can be embedded in HTML,
XHTML, or syndication formats like Atom or RSS. Although it lowers barriers to
entry, there are severe limitations to taking the microformat approach, which can
be seen in the kind of discussions that are taking place on wikis developing the geo
extensions to currently available microformats.
17
The contributors are not necessar-
ily GI domain experts, and they find themselves falling into a trap of scope creep. It
is often not clear what the specific purpose of these additional markup tags are, and
because they are limited to HTML, which is a document format not a data format,
there can be no explanation of what each markup tag means or how it should be
used. In other words, more semantics are needed and a more disciplined approach to
knowledge modeling.
The RDF community has answered this plea to lower barriers to entry by the
development of RDFa. RDFa is a W3C recommendation that enables the embed-
ding of RDF data within XHTML documents based on the microformat approach of
coopting the XHTML attributes. This means that RDF-compliant software agents
can extract this data from the XHTML attributes and understand the semantics in the
data. One such example of this type of search agent was Yahoo! Search Monkey.
18
Search Monkey used the RDFa metadata to enrich search result display and provided
a developer tool to extract data and build apps to display the Web site owner's own
custom applications. Although this closed in October 2010, the reasoning behind the
change was that SearchMonkey required developers to build a lightweight applica-
tion or service using the Search Monkey API, whereas Yahoo!'s new strategy is
to encourage Web developers to embed RDFa data directly in their Web sites, and
Yahoo!'s standard search can take advantage of this structured information directly,
rather than through an additional application that must be built separately for each
dataset. So far from the semantic experiment being regarded as a failure at Yahoo!,
this shows how it is moving into the mainstream. Another sign that the need to struc-
ture information is being more widely recognized is that schema.org, a collection
of markup recognized by the major search engines (Microsoft Bing, Yahoo!, and
Google), now includes a subset of RDFa.
A second trend in Semantic Web technology is an expansion of the infrastructure
required to support the Web of Data: development of the RDF databases (known as
“triple stores” as they are designed to hold triples). Several commercial ventures,
