Loading...
Please wait, while we are loading the content...
Similar Documents
Proteus: a frame-based nonmonotonic inference system (1988).
| Content Provider | CiteSeerX |
|---|---|
| Author | Russinoff, David M. |
| Abstract | Introduction Early artificial intelligence systems relied on first-order predicate logic as a language for representing domain knowledge. While this scheme is completely general and semantically clear, it has been found to be inadequate for organizing large knowledge bases and encoding complex objects. As an alternative, various frame-based languages have been employed. These languages are designed to support the natural representation of structured objects and taxonomies. They have proved to be well-suited for representing many useful relations, although they lack the general expressive power of the predicate calculus. Knowledge-based systems may also be classified according to inference methods. Most deductive systems may be characterized as either goaldirected (backward chaining) or data-directed (forward chaining). In a goal1 directed system, logical implications are encoded as rules that are used by the system to reduce goals to simpler subgoals. This allows knowle |
| File Format | |
| Publisher Date | 1988-01-01 |
| Access Restriction | Open |
| Subject Keyword | Frame-based Nonmonotonic Inference System Inference Method Deductive System Structured Object Complex Object Many Useful Relation Various Frame-based Language First-order Predicate Logic Natural Representation Domain Knowledge Goal1 Directed System Predicate Calculus Introduction Early Artificial Intelligence System Large Knowledge Base Knowledge-based System Logical Implication General Expressive Power |
| Content Type | Text |
