German NLP dataset for classification models in risk management

The dataset provides natural language data for training binary risk classification models in German language. It contains labelled text data for both risks and chances.

License: CC-BY-4.0, see https://choosealicense.com/licenses/cc-by-4.0/

Purpose and characteristics of the data set

• Goal: Provide natural language text data for training binary classification models in risk management applications
• Reason why: There aren't yet enough data sets publicly available that cover both German language and risk classes
• Each record of the text data set is labelled as follows: 1 = the issue is a risk, 0 = the issue is a chance

Current status of the data set

• Number of text data records labelled as "risk": 503
• Number of text data records labelled as "chance": 503