NIST envisions 'thinking machine'
- By Kathleen Hickey
- Jun 06, 2008
When is a word not a word? When it doesn't have a definition. When is a group not a group? When no one knows the members.
Paradoxes like these also must also be addressed in the technology world: Data classes must be created and relationships among data understood. Resolving such issues is the province of ontologists ' experts in word meaning and using appropriate words to build actionable machine commands . They have reached a concept agreement to create a technology system making it possible for programmers to build thinking machines that reason about complex problems.
The agreement was reached at the two-day Ontology Summit held during the National Institute of Standards and Technology's Interoperability Week last month. The summit was a joint initiative
between NIST, the Ontolog virtual community of practice and National Center for Ontological Research.
, said, 'The Ontology Summit established the critical set of requirements and ground rules needed before we can begin serious construction of the repository,' said Steve Ray, NIST's manufacturing systems integration chief who hosted the meeting. 'It will save enormous amounts of time and money and facilitate new, complex systems in all sectors for manufacturing control, supply chain management and even biomedical management systems.'
Ontology is important for sharing and reusing information in addition to exploring relationships among data. It is used in many arenas, including artificial intelligence, software engineering, biomedical informatics and information architecture.
Ontologies can be used to answer queries, publish reusable knowledge bases, export data to other systems, search across databases, and facilitate interoperability across multiple, heterogeneous systems and databases.
The agreement calls for an electronic, scalable, open ontology repository containing diverse collections of concepts, such as dictionaries, compendiums of medical terminology and product classifications. The system would enable distinguishable, computable, reusable, and sharable information, including data, documents and services.
Ontologists envision users having the ability to search and query across and within the different ontology sections of the repository. Its information would range from conceptual domains and specific disciplines of communities to technical schema, such as Resource Description Framework, part of the World Wide Web Consortium's specifications originally designed as a metadata data model; the Web Ontology Language, a family of knowledge representation languages for authoring ontologies; and the Common Logic Framework for a family of logic languages intended to facilitate the exchange and transmission of knowledge in computer-based systems, in addition to standard Internet languages such as Extensible Markup Language.
Kathleen Hickey is a freelance writer for GCN.