A formal associative model of logic programming and its abstract instruction set

Abstract

Associative computation is characterized by the intertwining of search by content and data parallel computation. This intertwining facilitates the integration of knowledge retrieval and data parallel computation. The paper describes a formal set of architecture independent rules for an associative model of logic programming, and an abstract instruction set. The model integrates knowledge retrieval, data parallel computation, and rule based reasoning within logic programming paradigm. An example of abstract instructions has been presented through the compilations of illustrative programs. Benchmark results have been presented. Benchmark result shows that tight integration of rule based reasoning and data parallel computation has reduced overhead on high performance supercomputers.<>