Research highlights from Fraunhofer IAIS, Dresden Branch of the Institute

Conversational AI

Intelligent dialog systems for various business sectors

At the Fraunhofer IAIS branch lab in Dresden, we develop intelligent dialog systems based on "Question Answering" that enable fast and efficient access to information via voice or text input. The results are complete solutions which are data protection-compliant and tailored to the particular needs of users in various application areas, for example in the automotive industry, medicine, production or tourism. We use modern AI technologies that combine Machine Learning and knowledge modelling. The special characteristic of this approach is that knowledge graphs lie in the background of the dialog systems, which combine generated knowledge from complex data sets with specific domain knowledge from experts and render the system suitable for complex enquiries. This concept of 'Informed Machine Learning' enables us to adapt the dialog systems to very specific technical vocabulary and the requirements of various application areas.

Explainable & Efficient AI: Center for Explainable and Efficient AI Technologies (CEE AI)

In addition, we are committed to bundling the AI competencies of the Dresden region, closely interlinking them with industry and ensuring rapid technology transfer into companies. To this end, we rely on close cooperation both with university research and with other Fraunhofer Institutes in the newly initiated "Center for Explainable and Efficient AI Technologies CEE AI". Under the joint management of TU Dresden (Prof. Dr. Frank Fitzek) and Fraunhofer IAIS (Prof. Dr. Jens Lehmann), research at CEE AI covers the entire spectrum from hardware support to device communication and transfer into practice. A special focus lies on the explainability of AI models ("Explainable AI") and the efficient implementation of AI technologies in hardware systems as well as Big Data & High Performance Computing ("Efficient AI").

• https://cee-ai.org/