ijcai 09

The conception of knowledge-based system, and expert systems in particular, is a major area in artificial intelligence (AI). Such systems are designed to achieve the performance of human experts in limited areas by exploiting a set of knowledge mainly obtained from these experts. Created around 1975, they had a definite impact on AI and a sometimes exaggerated media coverage. The term expert system disappears in favor of the broader concept of knowledge-based system (KBS) that is found in various fields of activities of AI. This concept is based on a separation between the knowledge needed to solve a problem and the reasoning mechanisms exploiting this knowledge (known as case control structures, interpreter, inference engines).

Artificial Intelligence

The artificial intelligence (AI) has significantly increased its economic importance in its ability to tackle new classes of problems (find statistics here), different from those treated with conventional computer. These problems relate to various human activities (perception, decision making, planning, diagnosis, interpretation of data, language comprehension, design) and have the common feature of requiring a reasonable exploitation of a large amount of knowledge, for the most specific of the area studied and learned from experts.

The conception of knowledge-based system that can perform functions of symbolic reasoning is currently an important research in the developments of AI. Such systems require particularly adequate representation of knowledge brought into play as well as effective mechanisms for exploitation of such knowledge or reasoning.

Early research in artificial intelligence (AI) date from the late fifties. Twenty years later, their first applications appear in the business. At the time, the AI ‌‌was mainly about using the power of computers to reason about words and proposals. Then they gave the name of the expert system (OS) for computer programs written by one or more experts in the form of rules and laws of operation. Integrating these two approaches, it was later an issue of knowledge-based systems (KBS).