Monday, April 23, 2018
01:30 PM - 04:45 PM
Many business metadata management undertakings stall when stakeholders from different organizations don't want to compromise over term names and definitions. (After all, "account" means something very different for Purchasing than it does for IT.) This renders the development of something as basic as a metadata dictionary untenable. The problem is magnified when it comes to modeling enterprise data. Data thesauri, while helpful, are not operable by computers and so may not really solve the problem they were intended to address. Semantic technology and ontologies provide metadata/information managers, data architects, and CDOs the means for accommodating stakeholders' different contextual realities while still improving the metadata's clarity, correctness, consistency, and completeness. This presentation will:
- Review some of the key challenges of metadata management.
- Introduce the fundamental building blocks of ontologies.
- Describe the "Unique Names Assumption" and why it is to be avoided.
- Explain the power of OWL equivalence properties.
- Provide examples of how semantics addresses these key challenges.
Dan Carey is an ontologist and data architect with 30 years of consulting experience, 25 of that spent designing databases, data models, and data strategies with major IT service firms. His current work involves using semantic technology to make client systems inter-operable in a data-centric manner. Previously, he has primarily supported government clients at the federal, state, and local levels, and has designed semantic technology products and data exchange standards to assist in military human resources management.