Generic Complex Team Analyses in Real-Time (PhD Thesis, ongoing)
In the course of the StreamTeam project we plan to improve the collaboration, coordination and monitoring of teams by enabling complex team analyses in real-time.
In this thesis we aim for solving the generic part of the StreamTeam project. That is, we plan to realize a generic and holistic infrastructure for performing complex team analyses in real-time. As one of our targets is to bridge the gap between powerful analysis software and domain experts, a main question is how to simplify the definition of real-time complex team analyses while still staying generic, i.e., scenario-independent. In this thesis, we plan to leverage graphical workflow abstractions and to automatically generate and deploy workflows consisting of powerful workers based on these abstract definitions.
This plan implies several subquestions that we have to answer:
- How to enable a simple graphical composition of an analsyis workflow in multiple layers?
- Which basic building blocks are necessary and how can they be parameterized the best?
- How to automatically translate a graphical definition into a workflow of an actual data stream processing system.
In order to utilize existing research results instead of reinventing the wheel we plan to base our infrastructure on an existing Open Source data stream processing system. Thus, our first task is to analyse the set of existing systems and identify the one that fits StreamTeam's demands the best.
Hasler Foundation (Cyber-Human Sytems program)