Dr. Lino Wehrheim

A joint project with
Mark Spoerer (University of Regensburg)
Manuel Burghardt (University of Leipzig)
Janos Borst (University of Regensburg)
Bernhard Liebl (University of Regensburg)

Part of the DFG priority programme 1859 “Experience and Expectation”

Description

Financial economists and psychologists agree that collective sentiment plays a crucial role in financial markets. In this project, we created data for a daily aspect-based index that captures the sentiment at the Berlin Stock Exchange from 1872 to 1930, a time when Berlin was the key financial market in Germany. This index is based on market reports published every trading day in the Berliner Börsen-Zeitung, which give a verbal description of the sentiment among market participants. Due to daily publication and the long observation period, our corpus consists of about 18,000 market reports. To derive sentiment values, we apply a combination of expert annotation and machine learning. Methdologically, this data covers two novel aspects: First, we focus on both historical and highly domain-specific language, a dual challenge that thus far has rarely been addressed. As there are many similar historical sources, such as reports by chambers of commerce, our solutions will be helpful for future research. Second, we address a characteristic but neglected feature of financial texts that might be relevant also in a broader sentiment analysis context. Particularly, we focus on aspect- and entity-based sentiment analysis.

The data is available at https://doi.org/10.57882/20230827-000.

Project–related Publications

Wehrheim, L., Borst-Graetz, J., Liebl, B., Burghardt, M., & Spoerer, M. (2025). More than a feeling. Introducing an NLP-based media sentiment index for the Berlin Stock Exchange, 1872–1930. Historical Methods: A Journal of Quantitative and Interdisciplinary History, 1–21. Doi: doi.org/10.1080/01615440.2025.2506427

Borst, J., L. Wehrheim, A. Niekler, and M. Burghardt. 2023. “An Evaluation of a Zero-Shot Approach to Aspect-Based Sentiment Classification in Historic German Stock Market Reports”. In: M. Ganzha, L. Maciaszek, M. Paprzycki, D. Ślęzak (eds.): Communication Papers of the 18th Conference on Computer Science and Intelligence Systems 37, pp. 51–60. Doi: 10.15439/2023F3725. 

Borst, J., J. Klähn, and M. Burghardt. 2023. “Death of the Dictionary? – The Rise of Zero-Shot Sentiment Classification”. Proceedings of the Computational Humanities Research Conference 2023. 3558, pp. 303–319. https://ceur-ws.org/Vol-3558/paper3130.pdf.

B. Liebl and M. Burghardt. 2021. “An Evaluation of DNN Architectures for Page Segmentation of Historical Newspapers”. In: 2020 25th International Conference on Pattern Recognition (ICPR), Milan, Italy, 2021, pp. 5153-5160. Doi: 10.1109/ICPR48806.2021.9412571.

Liebl, B. and M. Burghardt. 2020. “From Historical Newspapers to Machine-Readable Data: The Origami OCR Pipeline”. In: Proceedings of the 1st Workshop on Computational Humanities Research (CHR), pp. 351–373.

Wehrheim, L., J. Borst, and M. Burghardt. 2023. “Money Can’t Buy Love?” Creating a Historical Sentiment Index for the Berlin Stock Exchange, 1872–1930 (2023). In: Anne Baillot, Walter Scholger, Toma Tasovac et al. (eds.): Digital Humanities 2023: Book of Abstracts. Graz 2023. Doi 10.5281/zenodo.7961822.

Wehrheim, L., J. Borst, M- Burghardt, and A. Niekler. 2023. “„Auch heute war die Stimmung im Allgemeinen fest.“ Zero-Shot Klassifikation zur Bestimmung des Media Sentiment an der Berliner Börse zwischen 1872 und 1930”. In: Anna Busch & Peer Trilcke (eds.): DHd2023: Open Humanities, Open Culture, Belval/Trier, pp. 90–95. Doi 10.5281/zenodo.7688632.

Wehrheim, L., B. Liebl, and M. Burghardt. 2022. “Textual Data from Historical Newspaper Scans and its Challenges for “Guerilla-Projects””. In: Regensburg Economic and Social History (RESH) Discussion Paper Series 08, Discussion Paper, University Library Regensburg.

Data:

Wehrheim, L., J. Borst, B. Liebl, M. Burghardt, and M. Spoerer. 2023. “More than a Feeling: Dataset on Media Sentiment Regarding the Berlin Stock Exchange”. Doi 10.57882/20230827-000.