Knowledge Sharing Technology Project

Project Leader: Richard Fikes

Knowledge Sharing Technology is a research project in progress at the Stanford University Knowledge Systems Laboratory .

The objectives of this project are to develop technology and methodologies that will enable the sharing and reusing of knowledge bases. Many impediments, both technical and social must be overcome to achieve these goals. Two central problems which we are addressing in this project are the following:

Heterogeneous Representation Languages: There is no single knowledge representation that is best for all problems, nor is there likely to be one the choice of one form of knowledge representation over another can have a big impact on a system's performance. Thus, in many cases, sharing and reusing knowledge will involve translating from one representation to another. Tools are needed that can help automate the translation process.
Heterogeneous Ontologies: Even if the representation problems are resolved, it can still be difficult to combine two knowledge bases or establish effective communications between them. The absence of a shared vocabulary presents a further barrier, which could be removed through the development of shared sets of explicitly defined terminology, sometimes called ontologies. For such ontologies to be useful, the definitions provided must include declarative constraints that specify the semantics of the terms being defined, and the ontology must provide procedural methods that enforce those constraints when the terms are used in an application.

The project is being conducted in conjunction with and in support of the activities of the ARPA Knowledge-Sharing Effort. In particular, the objectives of our project are to develop technology and methodologies for:

Building ontologies in a form that is translatable into the specialized representation languages of multiple application system environments, and
Interchanging the reusable content of knowledge bases, including ontologies, among specialized representation languages.

The interlingua representation language being developed in the Knowledge-Sharing Effort is called a Knowledge Interchange Format (KIF). KIF is intended to be a language for communication (``literary publication'' of knowledge) and is designed to make the epistemological level content of a knowledge base clear to the reader, but not to support automated reasoning in that form.

The problems involved in interchanging knowledge bases are not yet well understood, and there is open debate as to whether a generally useful interlingua can be specified. In this project, we are attempting to inform that debate by developing tools for translating knowledge bases into and out of KIF (see Figure 1), and using those tools to conduct knowledge interchange experiments that will substantially test the viability and adequacy of KIF as an interlingua.

With respect to ontologies, we are focusing on what we consider to be a central barrier to the development and use of common ontologies, namely the lack of an integrated computational ontology development environment that provides practical tools for precisely specifying the meaning of terms in an ontology; incorporating, extending and integrating existing ontologies; and translating the resulting ontologies into the specialized representation languages of application systems. We are developing an ontology development environment called Ontolingua and are working with engineers who will use Ontolingua to identify and illustrate candidate high payoff extensions to the current PDES/STEP standards.

Figure 1: Translation into and out of the KIF interlingua

Current Funding: ARPA, NASA Ames, Lockheed AI Center

Project Staff (as of September 1, 1994)

Faculty: Richard Fikes

Research Associates: Adam Farquhar, Thomas Gruber

Scientific Programmer: James Rice

Visiting Scholar: Asuncion Gomez-Perez (Polytechnic Univ. Madrid)

Ph. D. Students: Sasa Buvac, Todd Neller, Greg Olsen, Wanda Pratt

Collaborators: Prof. J. van Baalen (Univ. Wyoming), Prof. Mark Cutkosky (Stanford Univ. Mech. Eng. Dept.)

References:

M. R. Genesereth, R. E. Fikes (Editors). Knowledge Interchange Format, Version 3.0 Reference Manual. Computer Science Department, Stanford University, Technical Report Logic-92-1, March 1992. Available on line.

T. R. Gruber. A Translation Approach to Portable Ontology Specifications. Knowledge Acquisition, 5(2):199-200, 1993. Available on line.

Bob Engelmore <rse@ksl.stanford.edu>