Options
2015
Master Thesis
Titel
Rule-based Process Orchestration
Titel Supplements
An Expert System for the Dynamic and Infrastructure-Independent Generation of Geospatial Processing Chains
Abstract
The constant improvements of remote-sensing technologies along with the emergence of new technologies to gather geo-related information has led to an exponential growth of geospatial data along with the range of potential applications, during the last decades. New technologies are required to target the challenges which arise not only from the variety but more importantly from the volume of so called Big Data. This thesis presents a modular and flexible software architecture, which aims to address these issues through the rule-based generation of dynamic process chains in respect to properties such as type and volume of data and an efficient utilisation of the available processing infrastructure. The use of an abstract workflow model enables the definition of processing steps in a concise and comprehensible manner without the need to consider the efficient execution. After the exploration and evaluation of related solutions a precise analysis of the context is carried out. This way, all interacting systems and data sources are identified and considered during the following design of the architecture. Furthermore, the requirements and quality goals such as high-availability and scalability are pointed out. To proof the feasibility of the technology-independent architecture, a prototypical implementation is presented which can be also used as a groundwork for further developments. Finally, real-world scenarios are used to show how such a system behaves in conditions simulating the practical use, which allows a discussion about strengths and weaknesses of this approach.
ThesisNote
Gießen, TH Mittelhessen, Master Thesis, 2015
Verlagsort
Gießen