Our research focuses on innovative solutions addressing information assets that are captured in large multimedia data sources. These information assets require new forms of mixed model-driven and data-driven algorithms for enabling enhanced decision making, insight discovery, behavioral analysis, coherence and anomaly detection, forecast, as well as process optimization. This comprises event, entity, and pattern recognition problems but also context modeling as well as their mutual employment for learning and adaption.
One of the highlights is the Competence Center for Deep Learning.
If you are interested in a Bachelor's or Master's thesis, please have a look at https://agd.informatik.uni-kl.de/en/ under „Thesis Topics“.
Semantic technologies make content accessible to computers. Their application creates a reliable communication basis for exchange in dynamic collaborations. Applied semantic technologies are the basis for transfer and translation between different modeling and conceptual worlds and a prerequisite for automated data and knowledge services. Formalized vocabularies and ontologies standardize information content and enable correct understanding between computer systems even in open, dynamically changing worlds.
In the topic area "Earth and Space Applications" techniques of machine learning are used to deepen our understanding of proccesses on our planet and its surrounding space environment. In particular, the team focuses on the use of earth observation data in combination with artificial intelligence for different application areas such as agriculture, urban planning, air pollution monitoring and flood mapping.
In the topic field of experience-based learning systems, we do research on AI systems that collect and learn from experiential knowledge for decision support or for solving complex problems. Such systems aim at understanding problems to be solved, abstract them appropriately, and propose solutions by accessing and reusing experiential knowledge.
The topic field of Immersive Quantified Learning uses artificial intelligence methods in combination with various sensor technologies to make human learning measurable and quantifiable. This makes it possible to provide learners with a learning experience that is individually adapted to their level of knowledge.
In the Cognitive Social Simulation team, we focus on modelling and simulation of humans or AI-systems as part of processes and systems. Humans and AI-systems as active elements enrich processes with their individual knowledge, specific expertise, individual decision-making, and their interaction with other process participants or AI systems. The method of cognitive social simulation is used for design of various applications in diverse domains ranging from simulation of means in the COVID-19 pandemic to Industry 4.0 systems.
In the topic area Multimedia Analysis and Data Mining (MADM), we are developing machine learning and data mining techniques to analyze and combine information from multi-modal inputs (e.g., combinations of image, audio, video, text, knowledge). In the focus of our investigations are especially also Deep Learning techniques, the efficiency of their training (GPU HPC), as well as their explainability (XAI).
In pattern recognition, we examine data of the most diverse nature for similarities, repetitions, and regularities in order to make them practically usable. Our fields of application include document analysis, medicine, the automotive industry, plant control and maintenance, and cyber communities.
The topic field Knowledge Work implements context-specific assistance systems to support information and knowledge workers in their daily work. This assistance is embedded in the user's personal knowledge space (Semantic Desktop) and the company's information space via a corporate memory infrastructure (CoMem).
Phone: +49 631 20575 1010
Deutsches Forschungszentrum für Künstliche Intelligenz GmbH (DFKI)
Smart Data & Knowledge Services
Trippstadter Str. 122