Dissertation/Thesis Abstract

Domänenspezifische Evaluation und Optimierung von Datenstandards und Infrastrukturen
by Schneider, Tobias, Dr.Nat., Universitaet Bayreuth (Germany), 2014, 470; 10692122
Abstract (Summary)

During the recent years and decades the amount of data collected in scientific projects has been growing dramatically. The demands on data persistence and infrastructures increased analogically. To cope with these challenges infrastructures for data exchange in specific domains have been built. Especially in the domain of life-sciences so called megascience platforms emerged which are responsible for the global data exchange. An essential attribute of these infrastructures are the supported data models and standards. To determine the appropriateness of a given data standard for a specific domain generic evaluation systems are needed. With the help of these evaluation systems weaknesses in existing standards and infrastructures can be identified and eliminated. The fundament of this evaluation is to structure the application domain into processes. In this thesis all areas of data standards and infrastructures will be analyzed separately behind the background of biodiversity and biodiversity informatics as application domains. The first part of this thesis covers data standards. The results in the areas of evaluation of data standards and infrastructures are generic and can be applied to other domains. As data standards are a fundamental part of an infrastructure, this subject will be discussed first. To accomplish this, a generic system for the evaluation of data standards is developed and applied on the most important data standards in the domain of biodiversity. Basis of this system is the analysis of completeness of a data standard with the 'Process Oriented Schema Evaluation' (POSE). It is shown that none of these data standards can fulfill the demands on a standard of the domain of biodiversity informatics completely. On basis of these findings PODSL-Biodiv is developed with the 'Process Oriented Data Schema Language' (PODSL) as a flexible data standard for the domain of Biodiversity informatics. The second part of this thesis is concerned with infrastructures. At first 'Infrastructure Evaluation Framework' (IEF) is developed as an evaluation system for infrastructures on the basis of processes and applied to important infrastructures in the domain of biodiversity informatics. It is shown, that the organization of an infrastructure has a large impact on the quality beside the pure technical capabilities. The 'Global Biodiversity Information Facility' (GBIF) network is identified as the most important infrastructure in the domain of biodiversity. Nonetheless even the GBIF network cannot fulfill the requirements on an infrastructure completely. To overcome these deficits, the concept of BDEI as a further development of GBIF is proposed.

Indexing (document details)
School: Universitaet Bayreuth (Germany)
School Location: Germany
Source: DAI-C 81/1(E), Dissertation Abstracts International
Subjects: Computer science, Systems science
Keywords: Data infrastructures
Publication Number: 10692122
ISBN: 9781392737538
Copyright © 2020 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy