DIFFICULTIES IN EXPERT SYSTEMS DEVELOPMENT 

Freeing resources for expert systems development is difficulties for two reasons. First, expert systems, like their parent field AI, represent a new technology. The second problem is that a number of companies and organizations are attempting to enter the arena of expert system development. Due to this scarcity, hiring a consultant to do expert systems work may cost more than can be justified.

Executed properly, the process of extracting the knowledge from the human expert to create a complete knowledge base cannot be rushed. Many expert system development projects are team efforts, which take extra time simply because of their nature. From another angle, time constraints can also pose a problem. Often, a time limit is placed on the completion of the expert system development project.

The basic of knowledge for the expert system is the human expert’s knowledge. If this knowledge is incorrect or incomplete, the same errors will be present in the knowledge base.

Expert systems now are not capable of acquiring knowledge directly. Expert systems now are not capable of dealing with knowledge that has a mixed representation. Current expert systems are not capable of refining their own knowledge bases.

Back to previous page