Formalized context

 

Contexts first appeared in declarative AI when they were presented as a possible solution to the problem of generality in 1987 in [McC87b]. Since then, contexts have found uses in various AI applications, including:

• managing large knowledge bases [Guh91],
• translating knowledge [BF95],
• modeling knowledge and belief [GSGF93],
• integrating data bases [FDFP95, FEFP95],
• planning [BM96],
• qualitative reasoning [Nay94], and
• common sense reasoning  [MB94].

The machinery of context provides a powerful platform for integrating bodies of information which were not originally intended to be used together. Today this view is becoming more and more accepted in the knowledge representation community. The examples are numerous, but of particular relevance to the HPKB project might be the fact that the Ontolingua system, being developed at Stanford's Knowledge System Lab, is planned to have a context mechanism implemented in order to facilitate coherent use of ontologies which were not originally intended to be used together.

• Basic Properties of the Formal Theory of Context
• Lifting and Translation
• Specializing Context
• Relations on Contexts
• Application to Natural Language Databases
• Planning Applications
• Barriers