IN51C-01 INVITED
DOLCE ROCKS: Integrating Foundational and Geoscience Ontologies--Preliminary Results for the Integration of Concepts from DOLCE, GeoSciML, and SWEET
Ontologies are being developed bottom-up in many geoscience domains to aid semantic-enabled computing. The contents of these ontologies are typically partitioned along domain boundaries, such as geology, geophsyics, hydrology, or are developed for specific data sets or processing needs. At the same time, very general foundational ontologies are being independently developed top-down to help facilitate integration of knowledge across such domains, and to provide homogeneity to the organization of knowledge within the domains. In this work we investigate the suitability of integrating the DOLCE foundational ontology with concepts from two prominent geoscience knowledge representations, GeoSciML and SWEET, to investigate the alignment of the concepts found within the foundational and domain representations. The geoscience concepts are partially mapped to each other and to those in the foundational ontology, via the subclass and other relations, resulting in an integrated OWL-based ontology called DOLCE ROCKS. These preliminary results demonstrate variable alignment between the foundational and domain concepts, and also between the domain concepts. Further work is required to ascertain the impact of this integrated ontology approach on broader geoscience ontology design, on the unification of domain ontologies, as well as their use within semantic-enabled geoscience applications.
IN51C-02 INVITED
Towards a core ontology for integrating ecological and environmental ontologies to enable improved data interoperability
Research in the ecological and environmental sciences increasingly relies on the integration of traditionally small, focused studies to form larger datasets for synthetic analyses. However, a broad range of data types, structures, and semantic subtleties occur in ecological data, making data discovery and integration a difficult and time-consuming task. Our work focuses on capturing the subtleties of scientific data through semantic annotations, which involve linking ecological data to concepts and relationships in domain-specific ontologies, thereby enabling more advanced forms of data discovery and integration. A variety of ontologies related to ecological data are actively being developed, ranging from low-level and highly focused vocabularies to high-level models and classifications. However, as the number of ontologies and their included terms increase, organizing these into a coherent framework useful for data annotation becomes increasingly complex (we note that similar issues have been recognized within the molecular biology and bioinformatics communities). We describe a core ontology model for semantic annotation that provides a structured approach for integrating the growing number of ecology-relevant ontologies. The ontology defines the notion of "scientific observation" as a unifying concept for capturing the basic semantics of ecological data. Observations are distinguished at the level of the entity (e.g., location, time, thing, concept), and characteristics of an entity (e.g., height, name, color) are measured (named or classified) as data. The ontology permits observations to be related via context (such as spatial or temporal containment), further supporting the discovery and automated comparison and alignment (e.g., merging) of heterogeneous data. The core ontology also defines a set of extension points that can be used to either directly build new domain ontologies (as extension ontologies), or to provide a common basis to which existing ontologies can be mapped. Thus, by enforcing a well-defined and high-level structure, the core ontology helps organize and integrate the growing number of ontologies, leading to better data discovery and integration techniques for semantically annotated data.
IN51C-03 INVITED
Pi in the Sky - Astrometry and the Virtual Observatory
A practicing stellar astrometrist relates trials, tribulations, and successes with present-day Virtual Observatory interfaces and databases, including the U.S. National Virtual Observatory site, SIMBAD/Aladin, and Google Sky. I briefly describe a recent astrometric result from the Hubble Space Telescope (measuring the mass of an extrasolar planet), aided and abetted by contributions from an existing Virtual Observatory. We look forward to a time when issues of database completeness, access speed, and *quality* of interaction have improved and further evolved, and present some suggested concrete requirements for next generation virtual observatories.
IN51C-04 INVITED
Leveraging Someone Else's Knowledge Based System
The Virtual ITM Observatory (VITMO) is a system that integrates data resources and other virtual observatories together to give the appearance of a seamless system. The level of integration varies depending on the capability of the individual data provider. Some sources are simply viewed as file systems from which the VITMO selects data. Others provide a wealth of individual services that must be selectively integrated. The highest level of integration occurs with peer systems that support ontologies or other knowledge-based representations of data resources. These systems allow VITMO to move the search from a low-level one where VITMO must maintain knowledge about all of the products, to a high level one where VITMO makes high-level requests of the other system and receives the desired responses without knowledge of low-level details. This latter approach is the one that VITMO is using in its integration with the Virtual Solar-Terrestrial Observatory (VSTO), an ontology based system. In this talk we will compare and contrast the experiences of integrating with a system providing a large variety of individual services and one that provides a seamless system-level mapping using ontologies. http://vitmo.jhuapl.edu
IN51C-05 INVITED
Explosion of Scientific Data and Virtual Observatories
The amount of scientific information is doubling every year. This exponential growth is fundamentally changing every aspect of the scientific process – the collection, analysis and dissemination of scientific information. Our traditional paradigm for scientific publishing assumes a linear world, where the number of journals and articles remains approximately constant. The talk presents the challenges of this new paradigm and shows examples of how some disciplines are trying to cope with the data avalanche. In astronomy, the Virtual Observatory is emerging as a way to do astronomy in the 21st century. Other disciplines are also in the process of creating their own Virtual Observatories, on every imaginable scale of the physical world. We will discuss how long this exponential growth can continue.
IN51C-06
The Application of Ontological Methods toward Coastal Restoration
At the fall 2006 AGU meeting the Information Technology and Systems Center at the University of Alabama in Huntsville debuted a tool for ontology based search and resource aggregation called Noesis. Since that time Noesis, with a new ontology for seagrass habitats in the Gulf of Mexico, has been utilized to support evaluations of potential seagrass restoration sites. The seagrass ontology was generated from a standard stressor conceptual model description for five species of seagrass common to the Northern Gulf of Mexico. Coupling the seagrass ontology with the existing atmospheric science ontology allowed scientists to locate and retrieve substantial information about the seagrass habitat as well as stressors that impact the habitat induced by climate change and short term atmospheric phenomena. A domain specific catalog of seagrass resources was constructed and an application ontology developed that mapped the keywords of the catalog to the combined (atmospheric and seagrass) ontologies of Noesis. Noesis uses domain ontologies to help the user scope the search queries to ensure that the search results are both accurate and complete. The domain ontologies guide the user to refine their search query and thereby reduce the user's burden of experimenting with different search strings. Semantics are captured by refining the query terms to cover synonyms, specializations, generalizations and related concepts. As a resource aggregator Noesis categorizes search results from different online resources such as education materials, publications, datasets, web search engines that might be of interest to the user. This presentation will give an overview of Noesis and describe how it has been applied to coastal restoration investigations.
IN51C-07 INVITED
Ontology and Knowledgebase of Fractures and Faults
Fractures and faults are related to many societal and industrial problems including oil and gas exploration and production, CO2 sequestration, and waste isolation. Therefore, an ontology focusing fractures and faults is desirable to facilitate a sound education and communication among this highly diverse community. We developed an ontology for this field. Some high level classes in our ontology include geological structure, deformation mechanism, and property or factor. Throughout our ontology, we emphasis the relationship among the classes, such as structures formed by mechanisms and properties effect the mechanism that will occur. At this stage, there are about 1,000 classes, referencing about 150 articles or textbook and supplemented by about 350 photographs, diagrams, and illustrations. With limited time and resources, we chose a simple application for our ontology - transforming to a knowledgebase made of a series of web pages. Each web page corresponds to one class in the ontology, having discussion, figures, links to subclass and related concepts, as well as references. We believe that our knowledgebase is a valuable resource for finding information about fractures and faults, to both practicing geologists and students who are interested in the related issues either in application or in education and training.