An object-centered representation for financial analysis

摘要

Our experiences with modeling knowledge for a computer-based problem solving consultant in the domain of financial marketing, using an object-centered knowledge representation facility are discussed. Financial marketing represents a class of problem solving activities that are fairly common in the marketing environments of companies that produce capital equipment. Solving these problems requires combining a wide latitude of skills with a vast repository of market data (past, current, and projected). We have been experimenting successfully in the past few years with a prototype system, FAME, that provides integrated interactive problem solving expertise for financial marketing. Our experiences with the modeling of FAME knowledge bases, and our strategy for the engineering and acquisition of knowledge bases required for large-scale expert systems are presented.