The first ‘Semantic Days’ conference was held under the auspices of the Norwegian oil and gas trade association, OLF, in Stavanger last April. Presenters included semantic luminaries Eric Miller (MIT/W3C) and Deborah McGuinness (Stanford), co-author of the W3C’s Ontology Web Language (OWL). The semantic web, the brainchild of the world wide web’s inventor Tim Berners-Lee, revolves around the idea that information on a web page should be presented in a way that will allow automated ‘discovery’ of business relevant information.
Semantic oil?
Following educational sessions on semantic web technologies and on their application in the public sector, the ‘semantic oil and energy track’ included presentations from Statoil, Hydro, Tieto Enator and the POSC Caesar Association. As Jon Atle Gulla, of the NTNU research organization explained, semantic search uses ‘ontologies’ to represent domain vocabulary, documents’ content and users’ information needs. The problem is that most documents today do not contain such information, and annotation of documents with semantic information is a ‘tedious and labor intensive task.’ Worse, such metadata is not even used by current search engines! Notwithstanding these issues, Gulla expects that in the future, ‘hidden’ ontologies will aid query interpretation and automate semantic indexing. Gulla then turned to the Norwegian semantic flagship project, the ambitious Integrated Information Platform (OITJ Vol. 10 N° 6). Gulla showed how ‘morpho syntactic’ search can use resources such as Schlumberger’s Oilfield Glossary to develop semantic relationships.
IIP
IIP project manager Svein Omdal described how XML data standards and the web ontology language (OWL) are being used to classify and retrieve information from ISO 15926 and other data models. The aim is to be able to visualize information from both operations and subsurface. Using an example of condition-based maintenance, Omdal described first generation technology leveraging domain specific XML schemas based on the POSC CAESAR reference data library. A second generation approach will leverage a common ontology ensuring that concepts are consistently defined across domains. ‘Reasoning software’ will combine data from several domains, automatically monitor and control equipment, order spare parts and prepare maintenance plans.
AKISO
Computas VP Roar Fjellheim described the ‘Active Knowledge System for Integrated Operations’ (AKISO) project. This includes a Statoil project to assure the exchange of information between the offshore rig, operations centers and expert communities of practice in the organization. Early database approaches failed because passive processes are not efficient. This led to the search for knowledge transfer using semantics and ‘active’ knowledge search. Over time a ‘knowledge resources map’ is built up. AKSIO deploys a drilling ontology in OWL (Protegé). Knowledge maps are represented with the resource description framework (RDF and Jena). Integration is performed in Microsoft SharePoint.
© Oil IT Journal - all rights reserved.