Search

Social media data is used for analytics, e.g., in science, authorities or the industry. Privacy is often considered a secondary problem. However, protecting the privacy of social media users is demanded by laws and ethics. In order to prevent subsequent abuse, theft or public exposure of collected datasets, privacy-aware data processing is crucial. This dissertation presents a concept to process social media data with social media user’s privacy in mind. It features a data storage concept based on the cardinality estimator HyperLogLog to store social media data, so that it is not possible to extract individual items from it, but only to estimate the cardinality of items within a certain set, plus running set operations over multiple sets to extend analytical ranges. Applying this method requires to define the scope of the result before even gathering the data. This prevents the data from being misused for other purposes at a later point in time and thus follows the privacy by design principles. This work further shows methods to increase privacy through the implementation of abstraction layers. An included case study demonstrates the presented methods to be suitable for application in the field.:1 Introduction 1.1 Problem 1.2 Research objectives 1.3 Document structure 2 Related work 2.1 The notion of privacy 2.2 Privacy by design 2.3 Differential privacy 2.4 Geoprivacy 2.5 Probabilistic Data Structures 3 Concept and methods 3.1 Collateral data 3.2 Disposable data 3.3 Cardinality estimation 3.4 Data precision 3.5 Extendability 3.6 Abstraction 3.7 Time consideration 4 Summary of publications 4.1 HyperLogLog Introduction 4.2 VOST Case Study 4.3 Real-time Streaming 4.4 Abstraction Layers 4.5 VGIscience Book Chapter 4.6 Supplementary Software Materials 5 Discussion 5.1 Prevent accidental data disclosure 5.2 Feasibility in the field 5.3 Adjustability for different use cases 5.4 Limitations of HLL 5.5 Security 5.6 Outlook and further research 6 Conclusion Appendix Referenc

Die Veröffentlichung wissenschaftlicher Daten als zusätzlicher Output neben den wissenschaftlichen Publikationen hat in den letzten Jahren immer mehr an Bedeutung gewonnen. Mittlerweile wird die Publikation der erstellten Daten sogar häufig als Bedingung an die Förderung wissenschaftlicher Projekte gestellt und die dafür benötigten Forschungsdateninfrastrukturen sind selbst Gegenstand der Forschung geworden (Kindling, Schirmbacher 2013; Bernard et al. 2013; Bill 2014). In der Fördermaßnahme „Nachhaltiges Landmanagement“ des Bundesministeriums für Bildung und Forschung (BMBF) wurde hierfür im Rahmen des Koordinationsprojektes GLUES (Global Assessment of Land Use Dynamics, Greenhouse Gas Emissions and Ecosystem Services) eine wissenschaftliche Geodateninfrastruktur (GDI) für den Austausch der wissenschaftlichen Daten aus Modellrechnungen und Simulationen aufgebaut. Im Beitrag werden die spezifischen Anforderungen einer solchen wissenschaftlichen GDI adressiert und entsprechende Lösungen gezeigt. Dies betrifft unter anderem die Inhalte der Metadaten für die Simulationsergebnisse und deren übersichtliche Repräsentation im Netz, die Beschreibung und Visualisierung der Entstehungsgeschichte von Modelldaten sowie webbasierte Recherche-, Analyse- und Visualisierungswerkzeuge.

Integrity constraints play a major role when the quality of spatial data is checked by automatic procedures. Nevertheless the possibilities of checking the internal consistency of the integrity constraints themselves are hardly researched jet. This work analyses the applicability of reasoning techniques like the composition of spatial relations and constraint satisfaction in networks of constraints to find conflicts and redundancies in sets of spatial semantic integrity constraints. These integrity rules specify relations among entity classes. Such relations must hold to assure that the data is fitting to the semantics intended by the data model. For spatial data many semantic integrity constraints are based on spatial properties described for example through qualitative topological or metric relations. Since integrity constraints are defined at the class level, the reasoning properties of these spatial relations can not directly be applied. Therefore a set of class relations has been defined which, combined with the instance relations, enables for the specification of integrity constraints and to reason about them.