Generally, how does Knowledge Synthesis Software work?Through an inferential network, AreteQ's "Q" software goes beyond keyword searches and knowledge analysis to find information of value implicit in text. Knowledge synthesis software allows a user to search not only for words or phrases directly stated in a document but also for concepts, words, or ideas that are implied--and it does it at a speed of up to 15 pages per second with an accuracy of 95%+.
In other words, not only does Q do knowledge analysis where it looks for words, phrases, concepts, and ideas--It also creates new knowledge by drawing inferential conclusions based on what it finds.
Specifically, how does Knowledge Synthesis Software work?Perhaps an example is the best place to start. Suppose someone wants to find out about spiders that live in the Great Smoky Mountains. He or she might simply type "spiders" and "Great Smoky Mountains" into a search engine. Of course, what would come back are articles or documents containing exactly those words, or perhaps slight variations of those words--an article containing "spider" instead of "spiders," for example.
However, if one used Q's Knowledge Synthesis Software, an expert on spiders and an expert on the Great Smoky Mountains would have prepared a knowledge base to support the search. In this knowledge base would be a variety of inferential relationships, including one that lists the various spiders that live in the Smoky Mountains, like "brown recluse", "black widow"," wolf spider", etc. Likewise, geographical names, like "Cades Cove", would be included. Even slang could be a part, like "cobweb", "eight-legged bug", and "attacob", or words that suggest the region, like 'hillbilly."
Thus, if someone entered "spiders" and "Great Smoky Mountains," Q would correctly identify an article entitled, "Brown Recluse Habitat in Cades Cove," even if the words "spiders" and "Smoky Mountains" never appeared in the article. It would even find "The Hillbilly's Old Timey Cobweb Remedies" because "Hillbilly" might imply or suggest the Smoky Mountains and "cobweb" certainly implies spiders.
What are knowledge engineers?Q employs a knowledge base created by knowledge engineers working with subject matter experts. These experts might be criminal investigators who want a program to rapidly search computer hard drives for evidence; they might be classification experts at a nuclear facility who want to search outgoing documents for potentially sensitive or classified information; they might be counter-intelligence agents trying to spot suspicious emails. Whoever they are, the experts provide a knowledge base or network of concepts that might be equivalent to each other or implied by each other, etc. The Q engine uses this knowledge base to search documents.
Are knowledge engineers necessary to create knowledge bases?Knowledge engineers are trained to turn a system of knowledge containing equivalencies, inferences, patterns, links, etc. into a knowledge base that the Q engine can use. They work with subject matter experts to create a custom knowledge base for each application.
Contact us! >>That depends on how complicated the knowledge base is and how much computer experience the subject matter experts might have! A special program called the "Knowledge Base Modifier" or KBM is bundled with the software. This program allows non-computer types to create new knowledge bases or maintain and update already created knowledge bases. It is usually faster and more economical in the long run to have AreteQ's knowledge engineers work with the subject matter experts in the creation of the first knowledge base and to train the experts in how to create and maintain knowledge bases.