Curated Reasoning by Formal Modeling of Provenance

• Main Author: Shaw, Kevin B
• Format: Others
• Published: ScholarWorks@UNO 2013
• Subjects: Provenance; Computer Science; Ontology; Enterprise Modeling; Curation; Reasoning; Other Computer Sciences
• Online Access: http://scholarworks.uno.edu/td/1782; http://scholarworks.uno.edu/cgi/viewcontent.cgi?article=2774&context=td

The core problem addressed in this research is the current lack of an ability to repurpose and curate scientific data among interdisciplinary scientists within a research enterprise environment. Explosive growth in sensor technology as well as the cost of collecting ocean data and airborne measurements has allowed for exponential increases in scientific data collection as well as substantial enterprise resources required for data collection. There is currently no framework for efficiently curating this scientific data for repurposing or intergenerational use. There are several reasons why this problem has eluded solution to date to include the competitive requirements for funding and publication, multiple vocabularies used among various scientific disciplines, the number of scientific disciplines and the variation among workflow processes, lack of a flexible framework to allow for diversity among vocabularies and data but a unifying approach to exploitation and a lack of affordable computing resources (mostly in past tense now). Addressing this lack of sharing scientific data among interdisciplinary scientists is an exceptionally challenging problem given the need for combination of various vocabularies, maintenance of associated scientific data provenance, requirement to minimize any additional workload being placed on originating data scientist project/time, protect publication/credit to reward scientific creativity and obtaining priority for a long-term goal such as scientific data curation for intergenerational, interdisciplinary scientific problem solving that likely offers the most potential for the highest impact discoveries in the future. This research approach focuses on the core technical problem of formally modeling interdisciplinary scientific data provenance as the enabling and missing component to demonstrate the potential of interdisciplinary scientific data repurposing. This research develops a framework to combine varying vocabularies in a formal manner that allows the provenance information to be used as a key for reasoning to allow manageable curation. The consequence of this research is that it has pioneered an approach of formally modeling provenance within an interdisciplinary research enterprise to demonstrate that intergenerational curation can be aided at the machine level to allow reasoning and repurposing to occur with minimal impact to data collectors and maximum impact to other scientists.