Publications   Software   Contact

Dr. Christoph Carl Kling
GESIS - Leibniz Institute for the Social Sciences

contact ät



Christoph Carl Kling. Probabilistic Models for Context in Social Media - Novel Approaches and Inference Schemes. Dissertation, 2016, summa cum laude (pdf)

Christoph Carl Kling, Lisa Posch, Arnim Bleier, Laura Dietz. Topic model tutorial: A basic introduction on latent Dirichlet allocation and extensions for web scientists. Tutorial, WebSci, 2016 (material)

Damien Fay, Hamed Haddadi, Michael C. Seto, Han Wang, Christoph Carl Kling. An exploration of fetish social networks and communities. NetSci-X, 2016 (arxiv)

Christoph Carl Kling, Jerome Kunegis, Heinrich Hartmann, Markus Strohmaier, and Steffen Staab. Voting Behaviour and Power in Online Democracy: A Study of LiquidFeedback in Germany's Pirate Party. ICWSM, 2015, Honorable Mention Award (pdf).

Christoph Carl Kling, Jerome Kunegis, Sergej Sizov, and Steffen Staab. Detecting non-gaussian geographical topics in tagged photo collections. WSDM, 2014 (pdf)

Christoph Carl Kling and Thomas Gottron. Detecting culture in coordinates: Cultural areas in social media. International Workshop on DETecting and Exploiting Cultural diversiTy on the Social Web, 2011 (pdf)

Christoph Carl Kling, Sergej Sizov, and Steffen Staab. Virtual field research with social media: A pilot case of biometeorology. WebSci, 2011 (pdf)

↑ up ↑


The implementations of my probabilistic models are free software licensed under GNU GPLv3.

Promoss topic modelling toolbox. This java project contains all my recent topic models I am developing since 2016, including the multi-context topic model described in my dissertation.

Multi-Dirichlet Process Geographical Topic Model (MGTM). My first implementation of a Multi-Dirichlet Process (MDP) topic model, which uses a collapsed Gibbs sampler. This model yields nice topics, but gets stuck in local optima. I still use this model if I want to detect a low number of human-interpretable topics. However, if you want to have perfect results and a faster inference, use the multi-context topic model of the Promoss toolbox.

↑ up ↑


You won't find me on social media.

Feel free to write me an email:

contact ät

Or write me a postcard:

Christoph Carl Kling
Unter Sachsenhausen 6-8
50667 Köln

↑ up ↑