Jour Fixe: Felix Hamborg and Franziska Weeber "Towards Semi-Automated Content Analysis: Using Deep Learning to Assist Manual Annotations"

The Zukunftskolleg invited everyone to our jour fixe led by Felix Hamborg and Franziska Weeber.

We invited you to our hybrid Jour fixe on Tuesday, 26 April.

Felix Hamborg (Associated Fellow / Dept. of Computer and Information Science) & Franziska Weeber (Dept. of History and Sociology) spoke about "Towards Semi-Automated Content Analysis: Using Deep Learning to Assist Manual Annotations".

Abstract:

Much research in the social sciences heavily relies on manual annotation, e.g., of textual documents such as party manifestos or news articles. And for a good reason, because only humans can reliably recognize subtle nuances in texts and read between the lines. However, these analyses are pretty costly and can take much time since they are conducted manually. This talk will give an overview of our previous and future research on making these analyses more efficient by using state-of-the-art natural language processing techniques, such as BERT. Specifically, we will summarize our prior work on the automated identification of media bias in news articles. Then, we will outline our current research project, in part funded by the Zukunftskolleg, to extend our approach to further use cases in social science research.