Positive feedback processes underlying the formation of expertise

Experts may be modeled as managing the inductive dynamics of knowledge acquisition in the knowledge processes of society. Who becomes an expert may be modeled as a random process under the influence of strong positive feedback loops in the social mechanisms giving access to knowledge. These models have implications for the design of expert systems. 1 The Nature of Expertise Expert systems add a new element to knowledge acquisition and transfer processes in human society. However, the nature and formation of expertise is not well understood. In taking part of the social process of knowledge transfer and embedding it in technology, it is desirable to have a more overt understanding of what an expert is and does. Use of the technology also raises questions about the possibly changing role of experts—what will experts do once their expertise is captured—are they accountable for the advice given by expert systems based on their expertise? Hawkins (1983) has abstracted from industrial experience in developing mineral exploration expert systems and proposed a model of human expertise relevant to expert systems. His model is summarized in figure 1: • The expert first elicits data about the problem from the client; • He or she develops a minimal model that accounts for the data provided; • He or she generates advice based on the model and feeds this back to the client; • The client may accept the advice, or query it and, possibly, the model; • The queries lead to further data elicitation, and repeat of the modeling/advice/query cycle. Thus, in Hawkins’ model, the client plays an active role in further developing the model by providing more data until he or she is satisfied with the model and consequent advice. Expert advice giving and taking is part of a cycle of negotiation around a process of model formation. This model of the expert-client interaction as the negotiation of a mutually acceptable model may be given formal foundations in mathematical modeling theory. Figure 2 shows a general hierarchical modeling process (Gaines 1987) based on Klir’s (1985) analysis of model formation: • At level one, constructs are those distinctions that the particular modeling system makes, a language for describing of the world; • At level two, data are descriptions of actual case histories in terms of the constructs, an account of experiences of the world; • At level three, hypotheses are the means of regenerating particular case histories from generalized accounts, rationalizations of the world (often called models); • At level four, analogies are similarities between differently generated generalized accounts, correspondences between models;