Information Technology Reference
In-Depth Information
majority of cases) segment it, identify picture primitives and also determine relations
between them. The classification proper (and also machine perception) consists in
recognising whether the given representation of input data belongs to the class of
images generated by the formal language defined by one of the grammars that can be
derived. Such grammars may be sequential, tree or graph grammars, and the recogni-
tion with their use takes place during the syntactic analysis conducted by the system
[25], [26].
Artificial intelligence techniques are used to support the correct reasoning based on
the data sets stored (collected). These techniques, apart from the simple recognition of
the data designated for analysis, can also extract significant semantic information
supporting the meaning interpretation of that data, that is its complete understanding.
This process applies only to cognitive information systems and is much more com-
plex than just the recognition, as the information flow in this case is clearly two-way.
In this model, the stream of empirical data contained in the subsystem whose job is to
record and analyse data interferes with the stream of
expectations
generated.
A certain
type of interference must occur between the stream of expectations generated by the
specified
hypothetical
meaning of the information and the stream of data obtained by
analysing the image currently under consideration. This interference means that some
coincidences (of expectations and features found in the data set) become more impor-
tant, while others (both consistent and inconsistent) lose importance. It leads to cogni-
tive resonance, which confirms one of the possible hypotheses (in the case of data
whose contents can be understood), or justifies a statement that there is a non-
removable inconsistency between the data currently perceived and all gnostic hy-
potheses which have understandable interpretations. The second case means that the
attempt at the automatic understanding has failed.
Cognitive information systems work due to cognitive resonance which is character-
istic for only these systems and distinguishes them from other intelligent information
systems. The use of such systems may be varied, as today's science provides broad
possibilities for them. However, the greatest opportunities for using cognitive infor-
mation systems are currently offered by medicine, which distinguishes more and more
disease units in the disease processes afflicting particular human organs, which units
are detected and recognised better and better. Medical images belong to the most
varied types of data and are characterised by extremely deep and significant (e.g. for
the patient's fate) interpretation of their meaning. Cognitive information systems are
also used in other fields, particularly in economics, where cognitive systems are
deployed to analyse strategic and financial figures of enterprises.
3 Cognitive Data Analysis and Interpretation Systems
The classification of cognitive data analysis systems is aimed at presenting the break-
down of cognitive systems and identifying areas of their application. In addition,
cognitive data analysis systems are also categorised, which means that selected
classes of decision-support systems are assigned labels, meanings and classifiers
which can be used to differentiate between cognitive systems. It is for this reason that
a classification of cognitive categorisation systems presented below has been intro-
duced. The following system classes have been distinguished among cognitive data
analysis systems:
Search WWH ::
Custom Search