Forschung
Unsere Forschung nutzt Data Science Methoden und Machine Learning zur Vorhersage menschlicher Entscheidungen auf digitalen Plattformen. Aktuelle Forschungsprojekte richten sich auf ein breites Spektrum an Forschungsfragen mit gesellschaftlicher und wirtschaftlicher Relevanz; einschließlich der Analyse sozialer Netzwerke, Finanzmärkte und methodischer Innovationen im Bereich der Text-Analyse.
- Research Focus
-
Forschungsfokus
-
Data Science for Business Applications
We use state-of-the-art quantitative methods to understand and predict the dissemination and economic impact of news, comments, and reviews in financial markets and electronic commerce. Furthermore, we engineer tools that allow decision-makers to replace gut decisions with data-driven practices.
Online Harms on Social Media
Social media is a fertile ground for misinformation and anti-social behavior, including online harassment, cyberbullying, and hate speech. Our research uses data science methods combined with large-scale datasets to better understand these phenomena and develop effective countermeasures.
Data Science Methods for Unstructured Online Data
Our research relies upon the ability to accurately process unstructured online data in various forms (e.g., text, images). For this purpose, we actively develop state-of-the-art data science methods (e.g., in the area of text mining and sentiment analysis) to derive actionable insights from unstructured online data.
-
- Laufende Drittmittelprojekte
-
Laufende Drittmittelprojekte
Community-Based Fact-Checking on Social Media, Deutsche Forschungsgemeinschaft (DFG), 2022 — 2025.
-
Rumor Diffusion on Social Media During the COVID-19 Pandemic, Deutsche Forschungsgemeinschaft (DFG), 2021 — 2025.
-
Einsatz neuer Daten in Kommunikationszusammenhängen auf Unternehmensebene (EnDiKaU), Land Hessen (Förderprogramm Distr@l), 2024 — 2027.
-
- R-Packages
-
R-Packages
-
Package: ReinforcementLearning
This package performs model-free reinforcement learning in R. The implementation enables the learning of an optimal policy based on sample sequences consisting of states, actions and rewards. In addition, it supplies multiple predefined reinforcement learning algorithms, such as experience replay.
ReinforcementLearning on CRAN
-
Package: SentimentAnalysis
This package performs a sentiment analysis of textual contents in R. The implementation utilizes various existing dictionaries, such as Harvard IV, or finance-specific dictionaries. Furthermore, it can also create customized dictionaries. The latter uses LASSO regularization as a statistical approach to select relevant terms based on an exogenous response variable.
SentimentAnalysis on CRAN
-
