Track: Data Analytics

Abstract

Collecting, processing and analyzing massive amounts of data embodies a steadily growing challenge for many companies. The ability to generate new knowledge from big data volumes nowadays represents a critical success factor. However, the capacity, capabilities and programming methods of currently existing IT-systems are unsuitable to handle big data processing and analysis tasks. An augmentation of traditional IT-systems with solutions that are sufficient for big-data storage and analysis is difficult and expensive. Furthermore existing IT-staff usually lacks the necessary expertise in installing and operating big data computing centers and applications. Therefore, it seems obvious to adopt a cloud approach. Several providers of cloud-based services meanwhile offer comprehensive big data solutions and appropriate analysis services.

A study conducted since 2014 examines the applicability of selected cloud platforms for data analysis tasks under special consideration of functionality, service models, costs, security, performance and scalability. The results deliver a catalogue of selection criteria and a comparison of analysis services for selected cloud platforms. The well-documented comparison is conducted on the basis of the use case “Analysis of failure data from wind energy plants”. It delivers templates for cloud-based big data analysis operation, but also depicts risks and limitations of cloud-based solutions.