TY - CHAP
T1 - Knowledge Base Design and Construction
T2 - From Prototyping to Refinement
AU - Kulikowski, Casimir A.
PY - 1989/1/1
Y1 - 1989/1/1
N2 - This chapter presents the design and construction of knowledge-based systems from prototyping to refinement. Designing and constructing an expert system depends on the type of problem solving that the system is trying to carry out. Despite the great variety of reasoning modalities employed in expert problem solving, most expert systems are designed to capture advice-giving or interpretive knowledge about how a problem is to be solved. The ultimate success of an expert system depends on acceptable proof that it is helping solve a problem more efficiently and effectively than available, though possibly nonexpert human counterparts, simple measurement of economic gain may suffice to demonstrate the advantages of the technological solution. However, because of the social repercussions of automation, and the ethical concerns about the correct application of codified human judgments, it is essential that the technical advances of the present generation of expert systems not obscure the great need for new insights into knowledge representation, reasoning, and the underlying semantics of knowledge bases. It is particularly important that a more systematic understanding of problem solving tasks be developed in relation to both surface models of compiled expertise and underlying models of scientific reasoning.
AB - This chapter presents the design and construction of knowledge-based systems from prototyping to refinement. Designing and constructing an expert system depends on the type of problem solving that the system is trying to carry out. Despite the great variety of reasoning modalities employed in expert problem solving, most expert systems are designed to capture advice-giving or interpretive knowledge about how a problem is to be solved. The ultimate success of an expert system depends on acceptable proof that it is helping solve a problem more efficiently and effectively than available, though possibly nonexpert human counterparts, simple measurement of economic gain may suffice to demonstrate the advantages of the technological solution. However, because of the social repercussions of automation, and the ethical concerns about the correct application of codified human judgments, it is essential that the technical advances of the present generation of expert systems not obscure the great need for new insights into knowledge representation, reasoning, and the underlying semantics of knowledge bases. It is particularly important that a more systematic understanding of problem solving tasks be developed in relation to both surface models of compiled expertise and underlying models of scientific reasoning.
UR - http://www.scopus.com/inward/record.url?scp=85012807108&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85012807108&partnerID=8YFLogxK
U2 - 10.1016/B978-0-444-87321-7.50011-9
DO - 10.1016/B978-0-444-87321-7.50011-9
M3 - Chapter
T3 - Studies in Computer Science and Artificial Intelligence
SP - 145
EP - 178
BT - Studies in Computer Science and Artificial Intelligence
ER -