Project: Ten Tweaks to Icicles Plots: A cool visualization for discussion quality in comments

Prof. Dr. Bernd Fröhlich
M.Sc. Dora Kiesel
M.Sc. Joshua Reibert

DegreeStudy ProgrammeExamination RegulationsECTS
M.Sc.Computer Science for 
Digital Media
PV18 and lower15
M.Sc.Computer Science for
Digital Media
M.Sc.Computer Science and Mediaall15
M.Sc.Human-Computer InteractionPV17 and lower15
M.Sc.Human-Computer InteractionPV1912/18


About 300 news articles are published online every minute, according to The GDELT Project ( Many platforms offer users the opportunity to discuss articles with others in the form of comments. The quality of these discussions has a significant influence on the opinions that both actively participating and consuming users form on the topic. For this reason, many social researchers, including the Chair of Empirical Methods of Communication Science headed by Prof. Dr. Ines Engelmann, are interested in finding out which factors at the level of the article and the news platform, as well as the comments themselves, have a lasting influence on the quality of the discussions. For this purpose, the chair collected 14.6k comments on 175 articles from 8 news platforms and manually tagged them with characteristics of the platform, the article and the comment itself.

Since the comments have a hierarchical structure due to interrelated responses, a representation of the data by an Icicle plot ( ) lends itself. The project will explore the Icicle plots possibilities to provide an interactive overview of the hierarchical structure and properties of the comments. Several aspects are challenging here, because both the number of comments per article varies greatly, as does the depth of the hierarchy that emerges. In order to meet these challenges and, in addition, to relate the quality characteristics of the comments to the characteristics of the articles and platforms, creative solutions are required, which we want to explore and realize using web-based technologies - first and foremost Javascript and HTML5.


Since we will be programming in HTML5 and Javascript, skills in these technologies are required. Ideally these skills cover also d3.js, a framework for realizing visualizations in Javascript.


The grade will be based on the active participation, the quality of the developed parts of the program, as well as the preparation and presentation of smaller talks and the final presentation.