150 likes | 290 Views
Categories of Vocabulary Compatibility. Dmitry Lenkov. Oracle. Interoperability. Interoperability layers. Delivery protocols Messaging protocols Content – syntax and semantics. Ability of two or more parties to. • Exchange information about metadata • Establish metadata equivalency
E N D
Categories of Vocabulary Compatibility Dmitry Lenkov Oracle
Interoperability • Interoperability layers Delivery protocols Messaging protocols Content – syntax and semantics • Ability of two or more parties to • Exchange information about metadata • Establish metadata equivalency • Exchange and understand metadata (or data about asset) • Exchange data (information) or provide service (car delivery)
Metadata interoperability • Fundamental for accessing data and services on the Internet • Relationships between multiple descriptions in terms of Interoperable (common semantics) vocabularies • Single query meaning over descriptions expressed in multiple descriptive formats (vocabularies) • Communities of interest based on a set of interoperable vocabularies
Data / Metadata exchange Source Destination Topic subscription no Subscription metadata generate validate valid? yes store confirm Info validate no Info metadata yes valid? generate validate valid? no no yes valid? validate request yes Info deliver Use and / or distribute
Interoperability Challenges • Explosion of Standards and Standard Organizations • Dozens of standard organizations and standards • No coordination • No consistency • High cost of multi-end, multi-standard interoperability • Development • Testing • Maintenance • Interoperability dependencies Vertical - on the deployment platform Horizontal - on multiple client/server/peer systems
Interoperability categories • Language level SQL, XML schema and query languages, ontologies Each language provides means to build vocabularies with Syntax to communicate data and Form queries about data • Compatibility of vocabularies Within one language – same XML schema Across languages – DM schema to XML Levels of compatibility – equality, equivalence, comparability
Entity / Data / Asset • Entity Generic term to items such as Cars, applications, books, scientific articles • Entity can be data Online book, document, Information item Digital asset • Entity can be non-digital asset Car, DVD with music or film • Entity has associated metadata in one or more vocabularies
Metadata / Data • Metadata is data about an asset or data It is a set of statements about an asset in one or more vocabularies If an asset is data (information), these vocabularies usually are different from vocabularies expressing data • Metadata can be of three kinds: Description – specifics of an asset or data Query – short term expression of interest Filter (subscription) – long term expression of interest • Distinction between description, query, and filter is purely functional
Vocabularies • Vocabulary Set of terms representing classes of entities, or instances Includes data types model, if any • Taxonomy Set of relations between classes of entities Relations - Subsumption, instantiation, equivalence • Ontology Is a vocabulary with non-empty taxonomy and a set of constraints in the associated constraint language
Data and Metadata Hierarchy describe data/metadata for Vocabulary Descriptions / Queries describe data/metadata for Vocabulary Descriptions / Queries describe Entity (asset)
Vocabulary Equivalence - Example • Example: Equal vocabularies (A & B) Same terms, same data type model, same taxonomy Is it enough to say that they are equal? – NO! • Evaluate vocabulary metadata Is the same vocabulary C used for A’s and B’s metadata? Yes, evaluate A’s and B’s metadata – same constraints, conversion (XSLT: A to B), set of synonyms, etc. - ? • No, different vocabularies - for A’s & B’s metadata Recursion – apply the process to A’s & B’s metadata vocabularies
Vocabulary Equivalence • Vocabularies A and B are equivalent if A’s and B’s metadata are specified in the same vocabulary C and A’s and B’s metadata specify same set of constraints and/or a conversion (XSLT A to B, taxonomy mapping, etc.) and/or a set of synonyms, etc. • else if for A’s and B’s metadata vocabularies C and D correspondingly are used and • Vocabularies C and D are equivalent and A’s and B’s metadata satisfy condition
Comparable vocabularies • Vocabularies with equivalent subsets How to deal with it Default values or instances Optional fields and partial evaluation • Examples Vocabulary evolution and versioning Independent creation for the same purpose and the same semantics
Mediation vs. Pear-to-pear • Mediation There is a set of (standard) vocabularies supported by a mediator Vocabularies of all communicating parties should be equivalent to one or more vocabularies in the mediator set • Pear-to-pear Vocabulary equivalence is on a pear-to-pear basis Greater effort for each pear Higher degree of flexibility, independence, and trust
Q & Q U E S T I O N S A N S W E R S A