An Ontology-Driven Model for the Efficient Use of Provenance Information
College: Liberal Arts and Sciences
Award year: 2006-2007
In this research we combine the gathering of provenance information with upcoming trends in research on semantic models, namely ontologies. We will analyze all measured provenance information to research and develop an optimal ontology-driven model for the most efficient utilization of the gathered provenance information. The proposed research builds on two main lines of work. First, it leverages NCSA current efforts on information gathering about decision processes using geospatial electronic records and medical images. Second, it adds on my previous research on ontologies of geo-spatial phenomena and imagery. NCSA provides a unique opportunity for doing this research because of its high performance computing capabilities to deal with massive amounts of data and CPU intensive processes, as well as the availability of the software to gather the provenance information. This proposed research cannot reach the optimal ontology-based model without having all NCSA resources. Therefore, the use of NCSA computer infrastructure and the joint work with NCSA's research staff will lead to enhanced results both for NCSA's research and for my own research on the use of ontologies for integration of geo-spatial information. This follows up on Bajcsy and Clutter's work, "Information Gathering about Decision Processes Using Geospatial Electronic Records". There, they claim that a "future direction for this research could be to embed provenance into a workflow engine [....] The provenance data are represented currently as triples with (subject, predicate and object). More elaborate taxonomy and ontology representations, as well mechanisms for interfacing the provenance meta-data would be considered in the future". In this research we will focus on multiple types of data and process provenance information collected by previous NCSA research using computer-centric and human-centric information gathering mechanisms. We will create a hierarchical organization of provenance information in the form of ontologies.