Information Technology Reference
In-Depth Information
A concept hierarchy is a graph whose nodes represent concepts and whose arcs
represent partial order relation between these concepts. In a concept hierarchy,
the meaning of a concept is built on a small number of simpler concepts, which
in turn is defined at a lower level using other concepts.
Concept hierarchies are used to express knowledge in concise and high-level
terms. As P.Witold [17] pointed out that granular computing is an information
processing pyramid, concept hierarchy tree just has the form of pyramid, where
more nodes at lower level and they are all specific concepts, less nodes at higher
level and they are abstract concepts.
In [1], the authors pointed out that the granulation process transforms the
semantics of the granulated entities. At the lowest level in a concept hierarchy,
basic concepts are feature values available from a data set. At a higher level, a
more complex concept is synthesized from lower level concepts (layered learning
for concept synthesis).
In [19], the authors stated “Concepts are not isolated into the human cog-
nitive system. They are immersed into a hierarchical structure that facilitates,
among others, classification tasks to the human cognitive system. This concep-
tual hierarchy expresses a binary relationship of inclusion defined by the following
criterions:
Inclusion criterion
: Each hierarchy node determines a domain included into do-
main of its father node. Each hierarchy node determines a domain that includes
every domain of its son nodes.
Generalisation-Specialisation criterion
: Every node in the hierarchy, has differ-
entiating properties that make it different from its father node, if it exists, and
from the others son nodes of its father, if they exist. ”
Concept hierarchies may be defined by discretizing or grouping data, that
is, discretizing numerical data into interval, and grouping categorical data into
a generalized abstract concept. A total or partial order can be defined among
groups of data.
For example, we can evaluate the ability of a person from multiview, such as
'education', 'vocation', 'income', etc. And, for each view, e.g., 'education', we
can regard it from multilevel by relying on our relevant prior knowledge. For
example, prior knowledge about 'education' can be organized as the following
concept hierarchy.
Education
High Low
Doctoral Postgraduate Undergraduate Others
student student
Fig. 1.
Concept hierarchy of 'Education'
