The Semantic e-Science Framework (SESF) (IN51B-1042) (tw.rpi/portal/SESF)

1. The Semantic e-Science Framework (SESF) (IN51B-1042) (tw.rpi.edu/portal/SESF) Peter Fox (pfox@cs.rpi.edu ), Deborah McGuinness(dlm@cs.rpi.edu), Rensselaer Polytechnic Institute, Troy, NY 12180. Funded by NSF/OCI award OCI-0943761. Non-specialist Use-Case(s): e.g. Access to Science Data from Virtual Observatories SESF builds upon previous work Abstract Virtual Solar-Terrestrial Observatory(VSTO) utilizes leading edge knowledge representation, query and reasoning techniques to support knowledge-enhanced search, data access, integration, and manipulation. It encodes term meanings and their inter-relationships in ontologies and uses these ontologies and associated inference engines to semantically enable the data services (NSF). Semantically-Enabled Science Data Integration (SESDI) project implemented data integration capabilities among three sub-disciplines; solar radiation, volcanic outgassing and atmospheric structure using extensions to existing modular ontolgies and used the VSTO data framework, while adding smart faceted search and semantic data registration tools (NASA). Semantic Provenance Capture in Data Ingest Systems (SPCDIS) has added explanation provenance capabilities to an observational data ingest pipeline for images of the Sun providing a set of tools to answer diverse end user questions such as ``Why does this image look bad? (NSF). • Goal to design and implement a configurable and extensible semantic eScience framework (SESF). • Configuration requires research into accommodating different levels of semantic expressivity and user requirements from use cases. • Extensibility is being achieved in a modular approach to the semantic encodings (i.e. ontologies) performed in community settings, i.e. an ontology framework into which specific applications all the way up to communities can extend the semantics for their needs. • Accommodating the rapid advances in semantic technologies and tools and the sustainable software path for the future (certain) technical advances • Generalization of the current data science interface, we will present plans for an upper-level interface suitable for use by clearinghouses, and/or educational portals, digital libraries, and other disciplines. • Acknowledgement: NSF/OCI Strategic Technologies for Cyberinfrastructure (STCI) Key Technology needs Semantic data framework architecture serving multiple communities • The ontology web language- OWL 2.0 and it’s profiles, esp. OWL 2 RL • A services language SAWSDL and OWL-S • Ontology editors – Protégé 4.0 and CMap • An ontology reasoner – Pellet/ FACT++ • A rule language – OWL 2 RL (RIF) and SWRL • A rule processor - Pellet 2.0, Jena, Jess • Semantic query – SPARQL – Triple store – Jena, Virtuoso, etc. • An explanation interlingua - PML - proof markup language • A suitable explanation infrastructure - inference web. • A collaborative portal infrastructure – Drupal 6/7 • Application-level semantic tools; faceted search, provenance explanation and browse Ontology Evaluation and Modularization leads to Packaging Concept maps of the upper level of the use case, translates and maps upper level (non-specialist) to mid-lower level terms (specialist) One of the key elements in this new framework is the ability to configure the framework with specific importing on required modules with the SESF ontology modules performing the ontology configuration Developing the mid-level semantics to achieve the closing of the knowledge gap between education and science, requires educational use cases • A • B Spectrum of Semantics and Knowledge Representations • Declarative - The what of knowledge representations - ontologies • VSTO used this form (with reasoning) almost exclusively • VSTO also used semantic query (filtering) with SPARQL • PML Is an ontology • Data integration needs this • Procedural - The how (the process) of knowledge representation (involves some degree of skill, which increases as a result of practice) - rules (and logic) • SESF needs this form, especially to mediate service choices and needs beyond the expressiveness of languages like OWL Frameworks v. Systems Faceted Search/ Provenance • Prior to 2005, we built systems, •  now frameworks • Rough definitions • Systems have very well-define entry and exit points. A user tends to know when they are using one. Options for extensions are limited and usually require engineering • Frameworks have many entry and use points. A user often does not know when they are using one. Extension points are part of the design • You don’t have to agree, this was our view