Tip:
Highlight text to annotate it
X
Since its inception, the Web has already witnessed two major growth phases.
Around 1995, authors began sharing on the network a huge amount of documents.
By 2005, social networks allowed users to interact and collaborate with each other,
as creators of user-generated content.
Documents, content and social profiles were related by generic hyperlinks, whose actual
meaning was to be guessed by the reader from the originating context.
A third wave, usually known as semantic or linked data web, is silently making its way
to major services and is likely to gain mainstream visibility in the next few years.
In the semantic web, entities and resources are connected to each other at a conceptual
level by a network of hyperlinks whose meaning is explicitly stated, effectively turning
the web into a global, actionable knowledge base.
In 2010, Google acquired Metaweb, a semantic web pioneer that had been working on Freebase,
a knowledge base comparable in goals and scope to Wikipedia.
Freebase provided a foundation for Google's Knowledge Graph, which is now leveraged in
the last release of Google's search algorithm.
The structured info-boxes that complements Google's search results are automatically
generated from the descriptions provided by the Knowledge Graph.
Descriptions are organized as machine-readable sets of tagged properties and are linked to
each other into a navigable network.
The real value of linked data descriptions is in the advanced, semantic search capabilities
they enable, where relevant entities are matched against conceptual models rather than textual
keywords.
Consider this straightforward marketing question: the web certainly contains the information
required to answer it, but getting hold of it may be tricky .
Search engines may help you locate a relevant document where the question was addressed
by someone else, but your chances may be slim.
Wikipedia, as the premiere collaborative fact checking source, may help you answer it by
yourself.
Car info-boxes do indeed contain all of the information you need, but Wikipedia describes
some thousands car models: scanning them manually in order to consolidate the bits into an Excel
file is likely to be time-consuming.
Luckily, the DBPedia project is automating this process since 2007, extracting properties
and links from Wikipedia info-boxes and consolidating them into a semantic knowledge base.
The DBPedia knowledge base may be queried using a standardized query language, providing
the required answer in a couple of minutes.
In the past 20 years, the evolution of intranet services closely tracked the evolution of
the open internet.
The document web was followed by enterprise document management systems.
The rise of the social networks inspired extensive intranet collaborative systems.
Our bet is that semantic web technologies will forerun the widespread adoption of enterprise
semantic knowledge bases.