Ontology Driven Integration Framework (ODIF)
Summary:
Knowledge Based Systems, Inc. (KBSI; College Station, TX), has developed an example-based search engine that delivers fast and relevant results that are highly tailored to the needs, interests, and individual profile of the user. MDA has provided Phase I and II SBIR contracts for KBSI’s technology, which could prove useful in sorting through information used in the areas of command and control, battle management, and communications (C2BMC). In 2007, the company’s technology, called JackalFish, became available commercially on a “limited trial basis.” The product can be applied to a variety of text-intensive industries, from publishing to law to government contracting.
Technology Description:
KBSI refers to its approach as an ontology-driven integration framework (ODIF), and the company has packaged its technology as a product called JackalFish (formerly known as AIOXFinder).
Traditional search engines—like those used for searching the World Wide Web—still require users to cull through search results to find the very specific piece of information that he or she is seeking. KBSI’s technology addresses the problem of excessive results by applying user profiles known as ontologies to its core search capabilities. Ontologies are software models or “representations of knowledge” allowing a software application to understand the semantics of—the “meaning” behind—a user’s search query.
An ontology, therefore, reflects a user’s knowledge, interest, or expertise in a particular topic area. An ontology includes not only a vocabulary of terms of interest to a particular user, but it also includes a related framework of logic that shows the relationships and semantics—and thus meaning—among the terms. Many ontologies exist for free in the public domain, and users can readily add them to KBSI’s JackalFish. Users of JackalFish also can use tools to create their own ontologies, defining the specific areas and topics of interest to them.
KBSI’s JackalFish applies text-mining and natural-language-analysis to a user’s queries and ontology to produce specific, highly focused, highly relevant search results. Compared with traditional, mainstream Web search engines, JackalFish is considered a more finely tuned search application that delivers very precise user-specific results. Also, in contrast to Web search engines, KBSI’s product is geared toward searches on closed networks—such as an organization’s intranet or corporatewide computer system. Beyond mainstream Web engines, many similar products that compete more directly with JackalFish use vocabularies—a list of terms—rather than full-blown ontologies.
An organization could set up JackalFish capabilities in about a day, according to KBSI. When the product is installed, it automatically indexes all documents on an organization’s network so that JackalFish can direct users easily to the documents. KBSI used an open-source tool called Lucene as the foundation for JackalFish’s indexing capabilities. In developing JackalFish, the company also used a free tool called WordNet, which organizes thousands of English words into sets of synonyms and definitions that are interlinked based on semantics and other word relationships.
From the desktop user’s perspective, JackalFish has “example-based” search capabilities—which sets it apart from competitors, according to the company. The document-centric approach means that a user can feed into JackalFish an entire paragraph of text that could be used as search criteria. That paragraph would then be cross-referenced with the JackalFish -indexed files in the organization’s network to generate a set of search results that are within the context of the particular paragraph.
JackalFish is considered a relatively light application. Users need only a standard PC with Microsoft Windows operating system. The company plans to add a “learning” capability to JackalFish —allowing the technology to analyze patterns in a user’s search history and to adjust future results based on that history.
MDA Origins:
KBSI developed the search technology through an MDA-sponsored project for an “ontology-driven integration framework” (ODIF) that could be used for knowledge sharing and communication in large military enterprises or in multisystem computer application environments. MDA originally funded KBSI’s ODIF work through Phase I and II SBIR contracts. For missile defense, the technology could prove useful in sorting through information used in command and control, battle management, and communications (C2BMC). ODIF technology from KBSI already has been installed at the MDA C2BMC Experimentation Lab, an environment where emerging technologies can be tried and then guided to the highest technology readiness level. In addition to MDA SBIR support, KBSI’s ODIF work also has been funded in part by the Missile Defense Agencys THAAD Program, the Office of Naval Research, the Air Force Research Laboratory, and NASA. The first test installation of the technology occurred in October 2006 at Cape Canaveral Air Station.
Spinoff Applications:
JackalFish should prove useful in any organization that deals with vast amounts of text. Law firms and publishing companies are obvious potential users. And organizations that work with design documents, requirements-based documents, or source code should find JackalFish especially helpful. Moreover, engineering or construction firms could put the tool to use for tasks such as assembling proposals when bidding on government contracts.
Commercialization:
KBSI continues to refine the ODIF-based JackalFish product to make it ready for broader government use and for widescale commercial release. The company has a patent pending on the technology, but KBSI officials are considering filing for patents that would cover certain aspects of its know-how—with the possibility of eventually licensing or selling technology outright to another software vendor. JackalFish now is available on a “limited trial basis,” according to the company, which continues to work on developing user guides and other documentation for the product, while also refining the user interface. Additionally, KBSI is exploring ways to get JackalFish easily into the hands of new customers, especially Federal users. The company also is considering avenues that would bring in more development money for JackalFish.
Company Profile:
KBSI was founded in 1988 and has been awarded numerous major research and development contracts that have helped it to develop some of its infrastructure methods and tools. Among the key projects has been an Information Integration for Concurrent Engineering (IICE) project, conducted for the U.S. Air Force. Following the project, KBSI’s research solutions were systematically transitioned to commercial software products that have been adopted by users worldwide.
KBSI’s work has led to software products in a broad spectrum of application areas, including artificial intelligence and expert systems, geometric reasoning, computer-aided design and manufacturing, manufacturing systems design and analysis, enterprise integration, process modeling, computer-aided software development, systems simulation, business process design and development, and total quality management.
KBSI is headquartered in College Station, TX, with two additional locations. The company employs close to 80 people.
Contact Information:
Dr. Perakath Benjamin
Knowledge Based Systems, Inc.
Tel: (979) 260-5274
email: pbenjamin@kbsi.com
web: www.kbsi.com
|