The task of the professorship is to develop and apply AI-based methods for the analysis of research data in humanities, in particular for the analysis of image, audio, audiovisual and 3D data, as they have emerged in recent years, either through the digitisation of analogue cultural assets, e.g. in museums and archives, or through more recent cultural practices, e.g. YouTube, Twitch, etc. The advanced analysis techniques are based on the creation of models with supervised or unsupervised procedures that map semantic knowledge in such a way that it can be used for individual analytical steps, e.g. entity recognition or the calculation of similarity relationships. Computational humanities thus face special tasks: The specific characteristics of data in humanities make their analysis in such contexts a very special challenge, since usually few datapoints are available or they are not evenly distributed. The terms of order used, e.g. for periods of time or stylistic currents, are subject to great fluctuations on the part of the observers and thus present learning processes with special tasks. The complex sign relations of the symbol worlds can only be grasped systematically in the long term with multimodal procedures (e.g. text-image-sound, film-3D etc.).