Johannes Maas

Johannes Maas

Bachelor Thesis Student


Under the supervision of Krishna Subramanian, I worked on my Bachelor's thesis "StatWire: Visual Flow-Based Programming for Statistical Analysis" with the following abstract:

We introduce StatWire, a tool for statistical analysis that combines the power of
statistical programming with the ease of use of visual programming. While existing
approaches already offer both statistical and visual programming and allow
the user to switch between them, StatWire provides the user with a simultaneous
integration of both modes. Because of this symbiotic integration, the user can write
code in a powerful statistical programming language, but is naturally encouraged
to modularize the code he produces in a visual structure. Splitting up the analysis
into modules allows the user to efficiently modify and reuse parts of it. Additionally,
StatWire provides a visual overview over the main steps in the analysis,
making it easier to understand.
This thesis contributes the artifact design for StatWire which is accompanied by a
prototype serving as a proof of concept. We explain the significance of StatWire by
presenting scenarios that showcase its benefits. To pave the way for future work,
we suggest research questions to explore. Other domains can also potentially benefit
from the advantages that StatWire brings to statistical analysis.

Created by jmaas. Last Modification: Wednesday 30 of August, 2017 18:16:56 by jmaas.

Media Computing Group at RWTH Aachen