Computational Social Science Winter Symposium

At the 2nd GESIS Computational Social Science Winter Symposium 2015 (Cologne, December 2nd/3rd), we presented the R-package polmineR with a Posterposter_css_preview. The package is designed to make working with large, linguistically annotated multi-purpose corpora fast and efficient. It is available at GitHub. Please note: There are some significant changes happening on the development branch. Most importantly, I am moving from data.frames to data.tables. Some changes will make it much easier to offer a documentation that can be digested. Please watch out for v0.6.0 of polmineR that is scheduled to be available in February.

Read More

CLARIN Annual Conference 2015

At the CLARIN Annual Conference 2015 (Wroclaw, September 15-18 2015), I gave an invited talk on “Interaction and dialogue with large-scale textual data: Parliamentary speeches of migrants and speeches about migration”. Corpus analysis entails going back and forth between quantitative and qualitative approaches to develop interpretations. One of the tools I used for supporting my argument was a three-dimensional visualisation of cooccurrence graphs of speeches on migration and integration affairs of speakers in the German Bundestag. The cooccurrence graphs are now available online, as a “normal version”, and a “anaglyph version” (get red-cyan goggles for the anaglyph effect). Use controls to zoom in/out and to rotate the graphs.

Read More