Business Analytics
Die Forschungsgruppe "Business Analytics" betreibt multidisziplinäre Forschung an der Schnittstelle von Informatik, Wirtschaftswissenschaften und Data Science. Unsere Forschungsaktivitäten sind auf ein breites Themenfeld mit gesellschaftlicher Relevanz ausgerichtet, einschließlich sozialer Netzwerke, Finanzmärkte und Systemen zur Entscheidungsunterstützung. Zu diesem Zweck entwickeln und nutzen wir Data Science Methoden und andere computergestützte Verfahren, mit dem Ziel die menschliche Entscheidungsfindung im digitalen Zeitalter besser zu verstehen.
Leitung |
||||
Bernhard Lutz, Dr. |
||||
Mitglieder der Forschungsgruppe |
||||
|
Research Focus: Text Mining
The availability of information forms the basis for human decision-making. We use modern algorithms like word embeddings and multi-instance learning to study different types of textual data like corporate disclosures, product reviews, and fake news. For instance, we aim to link the textual content of corporate disclosures to important variables of interest like fraud, earnings, or stock returns. In this vein, we also develop trading strategies based on financial documents. In addition, we aim to understand several aspects of product reviews like helpfulness and their influence on sales. Finally, we apply text mining to social media contents like real and fake news to understand why users fall for fake news.
Research Focus: NeuroIS (Information Systems and Neuroscience)
The subfield of NeuroIS within the IS discipline addresses research questions about human behavior using measurements from neuroscience like fMRI, EEG, ECG, and eye-tracking. We study phenomena in the context of fake news, product reviews, and recommender to explain how humans respond to different stimuli. For this purpose, we established a NeuroIS lab with a wide range of measurement devices.
Research Focus: Decision Support Systems based on Reinforcement Learning
We consider optimization problems from the field of operations research. Among others, we focus on scheduling problems in the automotive industry and investment decisions based on noisy predictions about future returns and risks. For this purpose, we develop novel methods based on reinforcement learning to learn optimization via "trial-and-error". In addition, we address current challenges of training a reinforcement policy directly in the real-world.