- From: Stefan Decker <stefan@db.stanford.edu>
- Date: Sun, 13 Jan 2002 02:59:24 -0800
- To: www-webont-wg@w3.org
Dear all, please find below the requirements part of the document in text form. All the best, Stefan --- Requirements: The following requirements were derived from the provided use-cases. * Simple and Efficient Inference System. o The ontology representation language must be simple and concise (small memory footprint) (use case 1, use case 4, use case 9), since small devices don't have the computational power to handle expensive inference tasks. This requirement may rule out certain expensive inference mechanisms. o Can the functionality be broken down into layers? (use case 9). o Clear definition of limits of the language (use case 9) (complexity, ) Discussion Point: Is the use of Description Logics too expensive for small devices? * Complex types. The ontology representation language must be able to handle complex types, e.g., dates. (use case 10). What is the expressive power of the data typing language? * An ontology needs a unique reference (use case 1 , use case 11). That is, it is necessary to identify an ontology. How is indicated that a class belongs to an ontology? * Ontologies may reference each other or may be distributed (use case 1). Ontologies need an import statement * Ontology Evolution. The ontology needs to provide language primitives, which enable an agent to identify how new terms were derived from old terms (use case 1). * Instance Data may be associated with a service, which provides the instance data (use case 2). Properties must support access preconditions - namely, the monitoring network should be prevented from writing, while allowing write by the control network. Property values have meta attributes describing data freshness, polling/event interval or freshness goal, type, acceptable value range(s), exception conditions. Administrative updates are access controlled and versioned allowing only authorized updates and rollback should an update not work as intended. Discussion Point: Where is the border between ontologies and services? Should services and ontologies be intertwined? * Constraints checking: Given a large amount of instance data, it should be possible to check if the instance data confers to a given ontology (e.g., for cardinality constraints) (use case 4).
Received on Sunday, 13 January 2002 05:59:35 UTC