6. Computer-Based Consultation System for Mineral Exploration ? Introduction

The search for mineral deposits is a difficult high-risk enterprise whose likelihood of success is influenced by a great many factors. Among the most important of these factors are the knowledge and judgment brought to bear on an exploration problem. If, for example, a panel of expert highly experienced geologists could be convened to examine every prospect, it is at least plausible that fewer deposits would be overlooked in the exploration process. Unfortunately, whether a project is undertaken by a private company or a government agency, it is not usually feasible to convene such a panel. Our purpose in this chapter is to describe an experimental system of computer programs called PROSPECTOR that is intended ultimately to provide expert consultation on problems of mineral exploration and resource evaluation. Ideally, PROSPECTOR would serve as a surrogate for a panel of expert economic geologists. However, because the cost of computation is very low and is projected to drop by a factor of about 100 over the next two decades, systems like PROSPECTOR could be made available to the exploration enterprise at large. When consulting about a prospect, the PROSPECTOR system first gives the field geologist an opportunity to tell the program about the most significant features of the site: the major rock types, minerals, and alteration products. The system then compares these observations with a number of stored models of various types of ore deposits and forms alternative hypotheses about the types of deposits that might be present. The system next engages the geologist in a dialogue in order to obtain additional data that would be useful in resolving the multiple hypotheses. The geologist can interrupt the system at any time to obtain clarifications of questions, volunteer additional data, request the reason a particular question has been asked, or obtain a general summary of the state of the consultation process.