Analyzing Digital Behavioral Data

Social media data and other digital behavioral data (DBD) are important for analyzing social science topics in digital societies and for understanding the evolvement of socio-technical systems. GESIS offers methodological insights on how computational methods support social science research and off-the-shelf tools for mining social entities, enriching data and disclosing social structures. With GESIS Notebooks we also provide an infrastructure for reproducible research and for sharing computational tools in this area. The "Total Error Framework for Digital Traces of Human Behavior on Online Platforms" (TED-On) is our first step in building a comprehensive framework for systematic error detection in the collection, processing, and analysis of digital behavioral data. 


Methodology, Framework 

GESIS aims at providing a comprehensive framework for systematic error detection in the collection, processing, and analysis of digital behavioral data. With focus on social media data, we developed the Total Error Framework for Digital Traces of Human Behavior on Online Platforms (TED-On).

Paper | Extended Paper |
MTE Talk | Slides | Tutorial@FAT

Topic Modelling


Our Topic Modelling Portal enables stochastic data analysis for web scientists and computational social scientists. The idea is to explain the fundamental mechanisms and ideas behind topic modelling. We provide instruments to detect latent topics in large text corpora while considering contextual information. 

Tutorial | Toolbox | Publication 

GESIS Notebooks

Virtual Research Infrastructure

Explore GESIS Notebooks (beta) – we are building an online
environment for web based large-scale data analysis with software suits for coding languages like R or Python.

The infrastructure will include services for application, publication, and archiving. 

GESIS Notebooks



The HypTrails Framework allows comparisons of hypotheses about sequential behavior – examples for this are, how websites are navigated or how persons move through cities.

Tutorial | Code | Paper | Paper |


API, Tool

Use the WikiWho Tool for 'social' text mining and analyze editing and revising transactions of Wikipedia entries across languages. Data can be downloaded as data set or obtained via an API.

WikiWho API | Data | WikiWho Wrapper | Report | Tutorial 

Open Science

We support and implement
Open Science.

Please visit our work
on GitHub.

Please find more relevant information and services