Submitted Master Thesis of Julius Elias

Julius Elias presents his master thesis on the topic "Crowd sourcing approaches for restructuring online discussions in a graph-based model".

Abstract

The rapid growth of the Internet, mobile computing and e-learning promotes a progressive shift towards digital communication. With Massive Open Online Courses (MOOCs) counting more than a million enrolled students, the complexity of their discussions are no longer manageable by linear structures. Drawbacks like the lack of scalability in the number of participants of linear structures motivate the approach to structure discussions in a graph-based model and visualize them similar to a mind map. The desire for structured content inspires the approach of finding relevant restructuring mechanisms. To offer the possibility of structuring existing discussions, seeding discussion data is mandatory. In this thesis, an import and a restructuring mechanism are presented. Imports from two popular online platforms, GitHub and Gitter, are implemented. The restructuring mechanism is implemented by defining tasks that represent atomic restructuring units and hence, small units of change to the discussion graph. A concept for synchronizing changes is introduced, which allows applying the restructuring results on the synchronized platform. With a two-staged study, problems in current discussion systems are first confirmed and then addressed with particular tasks. Both problems in discussion systems as well as a subset of implemented tasks are evaluated for a crowd-sourcing deployment. The results suggest that the tasks are applicable to be executed via crowd-sourcing.