Knowledge Systems, AI Laboratory
Computer Science Department
Stanford
University
We are working with Sandpiper Software, under a subcontract with Lockheed Martin Advanced Technology Laboratory, to develop a system for the representation and reuse of engineering design objects. Representative design objects include documents in various formats, 2D and 3D drawings, multimedia data, time-based simulation data, software modules and simulation software. The focus of this project is to improve the effectiveness of distributed collaborative engineering teams, working in a heterogeneous environment, in selection and reuse of design objects
The KSL is assisting Sandpiper Software in augmenting the RASSP reuse and data management system with the following capabilities: providing ontology tools and libraries for rapid formulation of new ontologies to enable access and integration of heterogeneous information sources; and function-based retrieval of design objects within an environment of multiple heterogeneous information sources.
Ontology-based technology for heterogenous data integration - KSL is developing, in collaboration with Sandpiper Software, methods for integrating ontologies with the RASSP Reuse data manager. This effort will enable the translation and utilization of the information models associated with heterogeneous data sources, hence enabling integration of these sources with the enterprise data management solution. The methodology will be demonstrated in conjunction with the Sandpiper Software Advanced Reuse management system.
For the past several years the KSL has been developing a computational environment, <a href=" http://www-ksl-svc.stanford.edu:5915/>Ontolingua</a>, which permits users to create, edit, browse and save ontologies over the World Wide Web. An ontology names and describes the entities that are assumed to exist in a domain and the predicates that are used to represent relationships among those entities. The ontology therefore provides a vocabulary for representing and communicating knowledge about the domain and a set of relationships that are assumed to hold among the terms of that vocabulary. The ontologies are presented to the user as frames, with the familiar notions of classes, slots, instances, and relations, along with documentation strings for human reference. Ontologies can be imported in and exported to the Interface Definition Language (IDL), among other languages. The Ontolingua ontology library and editor are publicly available, and there are currently more than 150 active users around the world. This project will further extend the capabilities of Ontolingua for creating rich object class hierarchies required for advanced knowledge sharing tools.
Function Based Retrieval - KSL is applying its previous work in developing a functional representation language to augment the design of an object browser with the ability to retrieve objects based on their functional specifications.