With Business Intelligence (BI) an organization collects and analyzes internal and external data in order to generate knowledge and provide decision support on strategic and operational levels. All this functionality is realized by employing an extensive variety of BI-algorithms that implement everything from simple ETL tasks to complex analytics. In order to work efficiently each of these algorithms is highly customized and tuned for a specific task. However, with the growing need to employ BI-techniques in more and more application scenarios, “hand-crafting” specific algorithms for every new task becomes time- and resource-intensive and thus represents a bottleneck. This topic tackles this issue, by first looking into generally valid concepts and constructs of algorithm design and BI-algorithms. Based on these insights, the goal is to design of toolbox of conceptual building blocks, from which arbitrary algorithms can be created fast and easy by means of combination. Furthermore an interaction concept is needed that enables users to effectively utilize said building blocks in different scenarios.

Main Advisor at Technische Universität Dresden (TUD)
Co‐advisor at Université Libre de Bruxelles (ULB)