{Begin SubSec Loops Knowledge Bases}
{Title Loops Knowledge Bases}
{Text

{Tag LOOPSenvironments}

Loops was created to support a design environment in which there are community knowledge bases that people share, and to which they can add incremental updates.  This section describes our goals for this facility, the concepts that we have employed, and scenarios for using knowledge bases in Loops.

We have chosen the term knowledge base instead of data base to emphasize two things: the kind of information being stored and constraints on the amount of information.  Loops will be used mainly for expert system applications where relatively modest amounts of information are used for guiding reasoning.  This information (i.e., knowledge) consists of inference rules and heuristics for guiding problem solving.  This is in contrast to potentially enormous files of facts, for example, social security records for California.  Reflecting this difference of scale, we have optimized the implementation to support fast access and updating to a smaller amount of information which is expected to fit in main memory for any one session.  For example, we maintain an index to the object information in computer memory.


{include UsingKBsub1}

{include UsingKBsub2}

{include KBguts}


}{End SubSec Loops Knowledge Bases}