Concept Trees: Building Dynamic Concepts from Semi-structured Data Using Nature-Inspired Methods - Complex System Modelling and Control Through Intelligent Soft Computations - page 230

law of thermodynamics implies, natural systems tend towards maximum entropy, or
a minimal and balanced energy state. While biological ones tend towards order as
well, the non-biological ones tend towards disorder. Cellular Automata are there-
fore closer to biological systems, where a cross-over is required if non-biological
systems are going to exhibit the same levels of intelligence. Although, something
like wave motion in the sea shows a steady and consistent behaviour until the wave
breaks. So the equations of wave motion can certainly work together in a consistent
manner. Even the snow

ake shows consistent behaviour for its growth stage, and so
on. So with the non-biological systems, a consistent energy input can be the con-
trolling mechanism that triggers speci

fl

c mechanics. If this is lost or changes, the
system can behave more chaotically and may have problems healing or fixing itself.

Biological systems might be driven by something more than the speci c mechanic,

which allows for another level of control. Consider the

re-join (self-heal?) capability of the concept tree later, in Sects. 4 and 5 .
The Gaia theory (Lovelock and Epton 1975 ) should probably be mentioned. As
stated in Wikipedia:

— possibly overriding —

The Gaia hypothesis, also known as Gaia theory or Gaia
principle, proposes that organisms interact with their inorganic surroundings on
Earth to form a self-regulating, complex system that contributes to maintaining the
conditions for life on the planet

'

'

. So the inorganic elements of the planet have a
direct effect on the evolution of the biological life. Maybe the inorganic mecha-
nisms suggested here are for smaller individual events, than the more gradual self-
regulation of a large global system. Although in that case, they could still be the
cause for global changes.

4 Concept Tree Examples

The examples provided in this section show how the concept trees can be built from
text sequences. They also describe some problems with the process and the pro-
posed solutions. With these examples, it is more important to understand the
general idea than consider them as covering every eventuality. Section 5 then tries
to give a more formal de

nition, based on these examples. To show how a tree is
created, consider the following piece of text: The black cat sat on the mat. The
black cat drank some milk. If punctuation and common words are removed, this can
result in the following two text sequences:

Black cat sat mat
Black cat drank milk

From this, as illustrated in Fig. 1 , a tree can be built with the following counts.
The base set of

' ' ' '

black cat can be extended by either the set sat mat or the set

' ' ' '

concept set has been found from a sort that starts
with these terms, where combining the two sets of terms then reinforces the base.
It also appears when constructing these trees that sets of counts should in fact
balance, unless additions with missing information are allowed. The counts for the

drank milk . The base black cat