Technology

Information retrieval on the scale of today's knowledge stores is a monumental task, and one which is significantly improved by adopting semantic retrieval capable of high performance.

Jarg's founder, Ken Baclawski, proposed a knowledge representation model, called KeyNet™, that unifies and extends many commonly used retrieval mechanisms. He also developed a distributed architecture and indexing algorithm for high-performance retrieval. This architecture fully exploits the expressiveness of the KeyNet model, and is embodied in all Jarg products, notably the Jarg Index Server and SKIP.

The KeyNet system is designed to express knowledge based on an ontology for a single subject area. It is a powerful way to mine complex textual documents such as research papers in a single discipline, but is also especially well suited to non-textual information objects, for example, scientific data files, satellite images and videotapes.

With Jarg indexing technology, information expressed as KeyNets can yield very high-performance retrieval from a corpus having up to several million information objects, at approximately the same level of performance as smaller corpora.

The algorithm is based on the vector space model for information retrieval. It differs from the usual vector space retrieval systems in using distributed hash tables rather than trees for indexing.