Im Zerrspiegel des Populismus

Eine computergestützte Analyse der Verlinkungspraxis von Bundestagsabgeordneten auf Twitter

Dieses Repository bietet einige Daten und Skripte zum Paper:

• von Nordheim, G. & Rieger, J. (2020). Im Zerrspiegel des Populismus - Eine computergestützte Analyse der Verlinkungspraxis von Bundestagsabgeordneten auf Twitter. Publizistik 65, pp. 403-424,
• siehe auch EJO.

Für Fragen und Anmerkungen verwendet bitte den issue tracker.

Distorted by populism - A computational analysis of German parliamentarians’ linking practices on Twitter

This repository provides some data and scripts related to the paper:

• von Nordheim, G. & Rieger, J. (2020). Distorted by Populism – A computational analysis of German parliamentarians’ linking practices on Twitter [Im Zerrspiegel des Populismus – Eine computergestützte Analyse der Verlinkungspraxis von Bundestagsabgeordneten auf Twitter]. Publizistik 65, pp. 403-424,
• see also EJO.

For bug reports, comments and questions please use the issue tracker.

Related Software

• tosca is used for managing and manipulating the text data to a structure requested by ldaPrototype.
• ldaPrototype is used to determine a prototype from a number of runs of Latent Dirichlet Allocation.
• longurl is used for expanding short urls.
• urltools is useful for extracting url cores from urls.
• spelling is used for identifying non-German texts.
• batchtools is used for calculating (prototypes of) LDAs on the High Performace Compute Cluster LiDO3.
• ineq is used to calcaulate Gini coefficients.
• ggparliament is used for visualization of the parliament.
• beanplot is used for visualize advanced boxplots.

Usage

Please note: For legal reasons the repository cannot provide all data. Please let us know if you feel that there is anything missing that we could add.

Due to the limited possibility to provide the raw data, all created plots and CSV files are collected in the folders Plots and Parteiverlinkungen. The scripts EDA.R (explorative data analysis), plotGini.R and plotLDA.R indicate the corresponding scripts. The script LDA.R shows how we calculated the different LDAs. All necessary data for the calculation are available. As an example we have stored the version with K = 30 Topics as result in the folder Modellieren.

The CSVs in Parteiverlinkungen refer, as shown in the paper (p. 413), to the cleaned data set that is the basis of the text analysis, i.e. they describe the data set generated by the scraper. Since scraping is subject to technical limitations (for example, no audiovisual content can be stored), the total number of links in the data set of all tweets may differ from the number of successfully scraped links listed there.