|
Chapter Summary
Most IS / IT applications use the computer's abilities for high speed calculation, data manipulation, communication and large scale data storage - this applies to all the applications discussed, thus far, in this section of the book.
There are, however, many tasks that require more than just calculation and data manipulation - they require judgement which is a part of intelligence. Computers are not intelligent but we can use their capabilities to simulate intelligence - this area of research and application is known as artificial intelligence (AI).
Research into AI has been extensive and wide ranging. The two main approaches to AI can be categorised as knowledge based systems (KBS) and artificial neural networks.
KBS are also referred to as expert systems. A KBS is set up with a database of 'knowledge' - 'facts' from the problem domain coded in some knowledge representation formalism, typically 'if - then' rules. The system can then be fed with details of a problem and will use its 'knowledge' to give expert advice. How a KBS is set up and how it operates is further explained in the chapter.
The alternative approach to AI is the neural network. This is a attempt to use the computer to simulate human thought processes. It is a newer process than KBS and still the subject of extensive research.
AI systems are used for a number of applications where conventional computing is not appropriate. Early KBS packages tended to concentrate on the field of medical diagnostics. More recent application areas have included in the fields of fault detection, financial advice and fraud detection.
|
