Knowledge Representation

In some respects, Artificial Intelligence (AI) is hardly a single field of study, covering as it does problems such as natural language understanding, computer vision, medical diagnosis, automatic planning, and learning. Yet despite this diversity, virtually all AI work is built on the same foundation-the belief that intelligent behavior can be exacted from a system with an explicit formal cache of knowledge and processing algorithms that extract conclusions from that knowledge base. We will explore this belief and some of its consequences, including some issues involved in building real systems to support adequate representations of knowledge and their associated reasoning procedures.