Concept Trees: Building Dynamic Concepts from Semi-structured Data Using Nature-Inspired Methods - Complex System Modelling and Control Through Intelligent Soft Computations - page 226

processing likes to link these up. This would mean that real-time information
retrieved from sensors, for example, could be combined with more knowledge-
intensive, but static information provided by the Internet, to answer a wider variety
of queries. A hierarchical structure is also appealing for reasons of organisation and
search ef

ciency, and so as has been suggested previously by other researchers
(Robinson and Indulska 2003 ), at least a shallow hierarchy would be useful. The
largest network of information that we have at the moment is of course the Internet.
This is composed of many individual Web sites that contain information by them-
selves. However, the only relation to other Web sites is through hyperlinks that are
typically created by human users. This is really the only way to try and combine the
information provided into a meaningful whole. To try and turn the Internet into a
network of knowledge, the Semantic Web has thus been invented. With the
Semantic Web, the programs that run on the Internet can describe themselves
through metadata, which will allow other programs to look them up and be able to
understand what they represent. Metadata is and provides extra

descriptive information about the contents of a document or piece of information. If
this information is available in a machine-readable format, then computer-to-com-
puter interaction will be enabled as well as the typical human-to-computer
interaction.
While the Internet is the main source for information, an evolving area is that of
mobile devices, including the Pervasive sensorised (Hansmann 2003 ) or Ubiquitous
computing (Green

' data about data '

eld 2006 ) environments. The mobile environment, by its very
nature, is much more dynamic. The Internet contains static Web pages that once
loaded will remain on a server, at a site from where they can be located. With
mobile networks, devices may be continually moving and so they may connect and
disconnect to a network at different locations. Ubiquitous computing is a model of
human-to-computer interaction in which information processing has been integrated
into everyday objects and activities. An example of this would be to embed sensors
into our clothes, to identify us when we went to a particular location. This dyna-
mism actually presents problems to a network that tries to organise through
experience. The experience-based organisation requires some level of consistency
to allow it to reliably build up the links, but if the structure constantly changes then
this consistency may be lost. However, the mobile devices may be peripheral to the
main knowledge content. They would be the clients that want to use the knowledge
rather than the knowledge providers. For example, in the case of people wearing
sensors, it would be the building that they entered that would learn from the sensor
information and provide the knowledge, not the people themselves. The sensors
would continually be bringing new information into the environment that would
need to be processed and integrated. The paper Encheva ( 2011 ) also includes the
ideas of concept stability and nesting, which are central to the whole problem. The
following sections describe how the laws of nature have helped with building these
complex structures.