Knowledge base definition

Definition 2 Knowledge-based systems are computer programs that encode symboHc knowledge about domains and tasks, and solve problems by manipulating this knowledge using quaHtative techniques. 

Definition 2 is phrased in terms of knowledge-based systems rather than expert systems. No reference is made to expert human problem solvers. Definition 2 captures the sense that the representation and manipulation of knowledge is the source of such a system s power, whether or not that knowledge is dkecdy eHcited from a human expert. 

What then is the drug designer to do, in order to achieve selectivity, in the face of overall similarity between receptor sites Two knowledge-based approaches have been used, and therefore come within our definition of drug design. Both approaches rely heavily on computational techniques, and on a principle that increased selectivity will parallel increased affinity for the receptor of interest. 

Besides differences in the functional form and reference state, from a more practical point of view, the knowledge-based potentials differ also with respect to scope of atom type definitions and the amount of structural data used for their derivation. The number of different atom types ranges from 17 in Drug-Score to 40 nonmetal atom types in BLEEP. In all cases, the Protein Data Bank (321) was the source of the solved crystal structures. For BEEEP 351 selected complexes were used, whereas the PMF function was extracted from 697 complexes, and DrugScore was derived using 1376 complexes. In the latter case, the data have been extracted from Relibase (322,323). 

Finally, the inference tool closes the loop by assisting in the inference of process definitions from process instances. The inference tool analyzes process instances and proposes definitions of task and relationship types. These definitions are stored in a knowledge base which may be loaded into Rational Rose. In this way, bottom-up evolution is supported. For a more detailed description of the inference tool, the reader is referred to [390]. 

The Definition and Design Phase involves decisions to confirm the suitability of the System Concept and to determine detailed functional requirements. It covers all aspects of designing the system and selecting a development environment including knowledge base creation, migration to delivery environments, and user interfaces. 

MEPAS is not an expert system, because all logic paths and input are fixed. However, MEPAS does contain elements of an expert knowledge-based system and is of interest in terms of a framework for evaluating environmental concerns. MEPAS is one of the few currently available fully integrated systems that accomplishes the interactive definition, organization, and assessment of complex environmental problems. 

The definition of the higher grade Knowledge Base structure is the primary goal during the Knowledge Elicitation phase. 

---