Abstract
The project BIAS: Mitigating Diversity Biases of AI in the Labor Market is afour-year project funded by the European commission and supported by the SwissState Secretariat for Education, Research and Innovation (SERI). As part of theproject, novel bias detection methods to identify societal bias in languagemodels and word embeddings in European languages are developed, with particularattention to linguistic and geographic particularities. This technical reportdescribes the overall architecture and components of the BIAS DetectionFramework. The code described in this technical report is available and will beupdated and expanded continuously with upcoming results from the BIAS project.The details about the datasets for the different languages are described incorresponding papers at scientific venues.