It has been rather silent on this blog for some time, but many things have happened. In the following, I summarize some of my outreach activities in the past months. I start with the public presentations and media coverage and then list the most important publications. As usual, the information and the PDFs are also available on the respective pages of this site: publications, presentations, media coverage.
In June, I gave a talk in Berlin at the Lange Nacht der Wissenschaften (“Long Night of the Sciences”) and presented our project Networks of Outrage to the public. There is also a video of the talk (~11 min.).
I also gave a few interviews to journalists, which contributed to the following texts:
On our Networks of Outrage project:
A slightly different version was published online:
On data journalism in general:
How data journalism can be implemented into newsrooms:
I co-authored three conference papers since the beginning of this year:
Abstract: Which news sources do supporters of populist islamophobic groups and their opponents rely on, and how are these sources related to each other? We explore these questions by studying the websites referenced in discussions sur- rounding Pegida, a right-wing populist movement based in Germany that is opposed to what its supporters regard as is- lamization, cultural marginalization and political correct- ness. We draw on a manual content analysis of the news sources and the stances of Twitter users, to then calculate the overlap of sources across audiences. Finally, we perform a cluster analysis of the resulting user groups, based on shared sources. Preferences by language, nationality, region and politics emerge, showing the distinction between differ- ent groups among the users. Our tentative findings have im- plications both for the study of mass media audiences through the lens of social media, and for research on the public sphere and its possible fragmentation in online dis- course. This contribution, which is the result of an interdis- ciplinary collaboration between communication scholars in Germany and journalists in Austria, is part of a larger ongo- ing effort to understand forms of online extremism through the analysis of social media data.
Abstract: The article describes the development of the Styrian Diversity Visualisation Project focusing on three main parts: the development of the prototype web app which enables users to explore their own situation in relation to others within the wider community, a corresponding data blog demonstrating the journalistic potential of data visualisation, and, finally, the integration into a course for students producing data-intensive stories. Furthermore, it highlights the project results as a prototypical example of a data journalistic piece.
Abstract: Today, journalists increasingly deal with complex, large, and heterogeneous datasets and, thus, face challenges in integration, wrangling, analysis, and reporting these data. Besides, the lack of money, time, and skills influence their journalistic work. Information visualization and visual analytics offer possibilities to support data journalists. This paper contributes to an overview of a possible characterization and abstraction of certain aspects of data-driven journalism in Austria. A case study was conducted based on the dataset of media transparency in Austria. We conducted four semi-structured interviews with Austrian data journalists, as well as an exploratory data analysis of the media transparency dataset. To categorize our findings we used Munzner ́s analytical framework and the Data-User-Task Design Triangle by Miksch and Aigner.
Abstract: Statistical open data is usually provided only in the form of spreadsheets or CSV files. The developers of open data apps must either restrict themselves to managable bite-sized chunks of data, which can be consumed (read, parsed, and held in memory) in one go, or must install and maintain their own data server which the app can query on demand. The Styrian Diversity Visualisation (in German ”Steirische Vielfalt Visualisiert” or SVV) project demonstrates the use of a dedicated data server (triple store) to host large amounts of statistical open data. The SVV web app queries the data server dynamically using SPARQL queries to obtain exactly the data required at that particular time, greatly simplifying its internal logic. There is no need to parse and store entire data sets in memory.
That’s it for now. I promise more frequent updates in the future.