ABSTRACT
Artificial intelligence.—The notion of inheritance in knowledge representation first began to develop in the field of semantic networks, where it was used for organizing information into inheritance hierarchies of the type sort-of or kind-of (→INFORMATION, REPRESENTATION, SEMANTIC NETWORK). Since then, inheritance has become a fundamental part of object representation and object-oriented programming (→OBJECT). In an inheritance-based system, concepts are classified hierarchically: the most general concepts dominate the more specialized ones (→CATEGORIZATION, CONCEPT). For example [vehicle] can be seen as a more general concept than [automobile] and as a more specialized concept than [physical object]. This type of classification is found in all large-scale knowledge-representation systems (such as Douglas Lenat and Ramanathan Guha’s CYC systems or Thomas Gruber’s Ontolingua) (→KNOWLEDGE BASE).
