Since 2011, I have been responsible for the Master-level lecture series "User Modelling and Personalisation". As can be observed from the slides (available online), I put a strong emphasis on the intuition behind statistical methods, predictive techniques, evaluation setups and design approaches. I believe that examples and discussions of advantages and drawbacks of several alternatives for different situations are more important than detailed, lengthy explanations of algorithms or statistical tests (for this, we use a wide range of hands-on exercises).
I use the same approach when supervising students or doctoral candidates. Within the context of their assigned or chosen research project, they have a lot of freedom in further specifying their research focus, as long as they can explain the rationale.
In the context of the Seminar "Advanced Topics in Web Science", we organise reading groups in which recently published papers are thoroughly read and discussed. This has proven to be a very engaging way to connect theory and lecture content to inspiring research outcomes or engaging systems and demonstrators.
Recently promoted Ph.D. students
Internal (daily supervisor)
Ricardo Kawase, Leibniz University Hannover: Building and exploiting context on the web (2014)
Sergey Chernov, Leibniz University Hannover, Germany: Using Contextual and Social Links in Information Retrieval (2012)
External (member of the examination board)
Charlie Abela, University of Malta: Behavior Mining for Personalised Desktop Tool-
Francesco Osborne, University of Turin: Propagating User Interests In Ontology-
Based User Models (2014)
Amjad W. Hawash, Sapienza University of Rome, Italy: Introducing Groups to an Annotation