INTRODUCTION

Over the past two decades the computer has been prescribed as the solution to all known human problems and has conspicuously failed to bring about a complete solution to any of them. Which brings us to Artificial Intelligence. Many attempts have been made to produce a neat and concise definition of the term. Probably the best of these is that Artificial Intelligence incorporates all of those programming technique which will, over the next decade, produce solutions to all problems which the computer has failed to solve over the past two. It is indeed, we are told, the universal panacea!Narrowing our sights from the universal to the oilfield scale, this paper will attempt to examine those problems of exploration data evaluation which are amenable to the computer and to Artificial Intelligence techniques of programming, The most widely known and discussed type of A. 1. program is the so-called Knowledge-Based Expert System, Unlike conventional, "algorithmic" programs, the Expert System does not reach its solution by following defined, formal procedures which can be guaranteed to produce a mathematically correct solution so long as a complete set of input data is provided, even though the input data may be erroneous and the output logically absurd. This is often referred "to as the GIGO factor (Garbage In = Garbage Out). Instead, the Expert System is a "heuristic" program which follows a deductive process using a Knowledge Base of established facts under the direction of rules of inference or good judgement. Like the human expert, upon which it is modeled, the Expert System cannot guarantee to always be totally correct. On the other hand, within the realm of expertise, the expert system (or human) can be relied to never be totally incorrect and to be able to supply some kind of an answer even when input data is incomplete orimprecise.

This content is only available via PDF.
You can access this article if you purchase or spend a download.