Nach positiver Absolvierung der Lehrveranstaltung sind Studierende in der Lage...

After successful completion of the course, students have basic knowledge of user interfaces, “human in the loop” systems, and artificial intelligence (supervised & unsupervised learning, reinforcement learning). Students further learn about various human-AI interaction domains (recommender systems, chatbots, intelligent text entry, explainable artificial intelligence, user modeling, personalized and adaptive user interfaces). After successful completion, students are able to apply AI techniques to their own (and potentially other novel) interaction scenarios, and in combination with the human-centered design process. 

The course conveys methodological knowledge about the design and implementation of AI systems supporting or cooperating with human users.

Conveyance of fundamental knowledge:

-        User interface basics and key terms

-        Basics of AI systems (supervised, unsupervised, and reinforcement learning)

-        Recommender systems (definitions, collaborative filtering, similarity measures)

-        Natural language processing (syntax, semantics, tokenization, normalization, stemming, corpus, chatbot interaction)

-        Gesture recognition (sequence classification, Markov property, gesture and pose from video)

-        Implicit interaction and physiological sensing

-        User Modeling (imitation learning, generative methods)

-        Adaptive User Interfaces (automated optimization towards human factor parameters)

-        Explainable AI (local & global interpretability, LIME, SHAP, automated rationale generation)

Conveyance and practice of abilities and skill

-        AI-adapted human-centered design process (data collection – model development – evaluation)

-        Tensorflow, Unity ML-Agents

-        User data collection

-        Parameter estimation and model choice appropriate for the scenario

-        Human-centered evaluation of human-AI interaction scenarios

-        Practical implementation of key scenarios (gesture recognition, recommendations, etc.)

Intelligent User Interfaces conveys methods and knowledge for the design of AI-supported human-machine interfaces. The course integrates theoretical and practical knowledge at the intersection of human-computer interaction, human factors, and artificial intelligence. 

The final compulsory proof of performance is given the final exam and small assignments. In groups of 2, students will be handed out 5 assignments covering different practical exercises related to the lecture content. Student groups will have to complete the work, document, and present their results.

The overall assessment is based on the following ECTS-Breakdown:

-        Lecture content: 25h

-        Exercise content: 25h

-        Group assignments: 25h

-