Humans and machines represent their knowledge in many ways. Efficient communication of knowledge, whether between humans or between humans and machines, relies on an understanding of knowledge representation.
Machines typically rely on processing speed, their ability to search through troves of preselected data, to reach conclusions. Humans, on the other hand, often draw conclusions from very small data samples, drawing comparisons from many loosely related ideas. IHMC researchers are developing systems and tools for encoding everyday knowledge and interconnectivity in forms suitable for machine inference. They analyze the semantics of notations, such as mathematical diagrams and maps, and the use of formal notations to encode intuitive meanings, particularly in a social context involving human and machine agents. New languages also improve the machine’s ability to deal with imprecise information and ambiguity.
IHMC researchers are also collaborating with colleagues around the world to develop the standards and notations for the Semantic Web. The Semantic Web is an effort to draw on the existing Internet to provide richer content and meaning, interlacing the information between pages in a new and dynamic way. These standards are designed to work with all of the existing web pages, so nothing will have to be scrapped. IHMC’s Pat Hayes played a pivotal role in the development of several of the key technologies, including the Resource Description Framework (RDF), the Web Ontology Language (OWL) and SPARQL, the first web query language. RDF and OWL provide the framework for asset management and the sharing and reuse of data on the web. The initial framework was introduced in 2004, and it is increasingly deployed in commercial and open-source projects.
In the Semantic Web, the definitions of concepts and the relationships between them are encoded in binary linkages (RDF graphs). IHMC researchers noted the similarities between this notation and concept maps and have adapted CmapTools software for use in generating and viewing Semantic Web pages. The resulting system, called COE, has been widely adopted for displaying Semantic Web knowledge. IHMC researchers are active in the development of standard “ontologies” – bodies of formalized knowledge – for time, agency, action, geospatial information, regulation and other topics of wide application.
IHMC’s Hayes is also the prime author of the new ISO Common Logic standard, a very expressive first-order logic that has been deployed in projects aimed at improving homeland security operations, resulting in a new extended self-describing logic called IKL. IKL is being used as a standard knowledge representation formalism for a number of advanced national security applications. Future work at IHMC will extend COE to IKL and create an experimental IKL reasoning engine.