Ongoing Projects
Inflation narratives in large text corpora
(funded within the framework of the Cross-Disciplinary Labs of the House of Computing and Data Science (HCDS))
The Cross-Disciplinary Labs of the House of Computing and Data Science (HCDS) promote interdisciplinar[FU1] y research at the interface between digital methods and subject sciences. The project "Inflation narratives in large text corpora" is designed for three years and is led by Prof. Dr. Ulrich Fritsche and Prof. Dr. Ricardo Usbeck (Semantic Systems Research Group, Department of Computer Science).
The project uses an interdisciplinary approach as a common ground to analyze narratives in textual data. The aim of the project is to train inflation-specific language models on the basis of large German- and English-language corpora. Events are extracted from these language models and linked and displayed in the form of knowledge graphs. The knowledge graph enables the identification of precise and meaningful patterns and correlations that also allow causal statements. This research strand is supported by a large online survey to determine causal narratives among households in Germany. In a further application part, the effect of identified inflation narratives in media discourses on aggregate expectations and macroeconomic variables is analyzed.
- Contact at the chair:
Prof. Dr. Ulrich Fritsche
Max Weinig - Contact at the Department of Computer Science:
Prof. Dr. Ricardo Usbeck