background image
243
Introducing intelligence to the Net: the semantic web
The current problem with web content is that it is designed for humans to read but not
for machines to automatically process the semantics.The concept of the semantic web aims
at making web content machine-readable, which is a basic requirement for automated
services."The Semantic Web will bring structure to the meaningful content of Web pages,
creating an environment where software agents roaming from page to page can readily
carry out sophisticated tasks for users." (Berners-Lee, Hendler, Lassila, 2001)
Today's Internet relies on traditional knowledge models that are based on centralised
models where everybody shares the same understanding of words and their meaning.They
represent closed systems, where users can ask only a limited set of questions for the com-
puter to answer reliably. In addition, traditional knowledge systems rely on rules that deter-
mine inference, yet those rules only apply within this closed system and cannot be
transferred to other systems.
The semantic web, in contrast, tries to make knowledge systems more versatile by
opening up the closed structures.This happens, however, at the expense that some questions
and paradoxes remain unanswered. Hence, as Tim Berners-Lee explains, the challenges of
building the semantic web are twofold:
first, to provide a language that expresses both data and rules for reasoning, and
second, to create rules that can be exported to other systems.
"Adding logic to the web the means to use rules to make inferences, choose courses of
action and answer questions is the task before the semantic web community in the
moment." (Berners-Lee, Hendler, Lassila, 2001)
At the heart of the semantic web are personal agents, i.e. software programmes that will
allow their users to automatically collect content from diverse resources on the web, process
this information and exchange the results with other programmes. A typical process will
involve the creation of a "value chain" where content from different resources will be
assembled in single steps, with each step adding value to the final "information product"
that answers the user's request. In the future, the effectiveness of these personal agents will
increase tremendously as new machine-readable web content and automated services
become available.
Semantic web applications in cultural heritage
In the cultural heritage domain, one example of a semantic web application could be
an agent system/avatar system to exchange information on cultural heritage events. For
example, knowing that I need to go on a business trip to Berlin, my personal agent that
knows my preference for modern art, would automatically contact a cultural service
provider to obtain information about modern art events and exhibitions in Berlin during
the time of my stay.The agent would report back the event dates and opening hours and
after my confirmation, automatically book tickets.The scenario could be developed further,
involving paying the tickets, sending them to the hotel where I am staying during my trip,
etc.
However, for such systems to work and be accepted by users, a secure and protected data
exchange is of utmost importance.This means that all the personal information remains
with one's personal agent, while the service providers that facilitate the information
exchange do not have access to any of this personal data. Based on open standards, the
personal agent should also work independently of any user device and be "portable",
meaning that it could "travel" between different end user tools, be it a personal
IX TECHNOLOGY