Information Technology Reference
In-Depth Information
The circle of artificial intelligence researchers has frequently been criticised in
publications by Hubert Dreyfus, John Seale, M. Kickhard and Loren Terveen [16] and
many other scientists who pointed out that further manipulations of symbols are not
sufficient to understand language and are much less suitable for describing the correct
operation of human perception, understanding, behaviours and action.
Nowadays it is noted that to assess the phenomenon taking place, one can refer to
the subjective feelings, religious experiences and the free execution of calculations by
thought processes inside neurons. There is hope that the brain function mechanism
will correctly, suitably and very precisely explain computational theories, although
scaled computations of a high difficulty level are nothing else than what has been
implemented according to a certain data set.
Today, scientists constructing intelligent information systems pin their hopes on
the development of laboratory procedures for their further modification. Around the
world, there are whole laboratories dedicated to this subject, which compete with one
another to find new, better and better methods of implementing intelligence in infor-
mation systems, as they are competing for markets to sell those systems in, and their
future depends on the innovation of the solutions applied.
Intelligent information systems became the foundation for building cognitive sys-
tems whose purpose is not just the simple analysis of data consisting in its recording,
processing and interpreting, but mainly an analysis by understanding and reasoning
about the semantic contents of the processed data.
Every information system which analyses a selected image or piece of information
using certain features characteristic for it keeps in its database some knowledge -
indispensable for executing the correct analysis - which forms the basis for generat-
ing the system's expectations as to all stages of the interpretation it conducts. As a
result of combining certain features found in the analysed image with the expectations
- generated using the knowledge - about the analysed semantic contents of the image
or information, cognitive resonance occurs [17]-[24], [27]-[33]. In addition, cognitive
information systems use methods which determine semantic/structural reasoning
techniques serving to match patterns [28]. Consequently, during this analysis the
analysed structure of the image (or another form of data) is compared to the structure
of the image which serves as such a pattern. This comparison is conducted using
strings of derivation rules which enable the pattern to be generated unambiguously.
These rules, sometimes referred to as productions, are established in a specially de-
rived grammar, which in turn defines a certain formal language, called an image lan-
guage. An image (a piece of information) thus recognised is assigned to the class to
which the pattern that represents it belongs.
Cognitive analysis used in cognitive information systems very frequently takes ad-
vantage of a syntactic approach which employs functional blocks to analyse the
meaning and interpret the image [28]. The input image is subjected to pre-processing,
which includes image coding with the terminal components of the introduced lan-
guage, approximating the shapes of analysed objects, as well as filtering and pre-
processing the input image.
The completion of these stages represents the image anew in the form of hierarchi-
cal structures of a semantic tree and subsequent steps by which this representation is
derived from the initial symbol of the grammar [28]. At the stage of data pre-
processing, an intelligent cognitive recognition system must (in an overwhelming
Search WWH ::
Custom Search