I am scientific contractual employee at the Institute of Psychology at University of Graz, Austria. As a computer scientist, I work mainly on administrative tasks but recently I found the programming of R Shiny apps to be very interesting. These apps, as my own research in general, deal with knowledge space theory, a psychological framework for structuring knowledge domains based on prerequisite relationships.

Knowledge Space Theory

Knowledge Space Theory was founded by Jean-Paul Doignon and Jean-Claude Falmagne in 1985. It is a great framework for structuring domains of knowledge based on prerequisite relationships. These knowledge structures can then be applied, e.g., for adaptive testing or personalised training.

Within knowledge space theory, my focus is on the adaptive assessment of knowledge. Already during the assessment, previously received answers (and their correctness) are used to infer whether or not other questions can be answered by the individual. Thus, the number of questions to be asked during the assessment can be reduced significantly.

The respective procedures are computationally quite expensive. Therefore, I am investigating more parsimonious alternatives.

R - Statistical Programming

R is a free software and programming language for statistics. There exist thousands of packages extending R's functionality.

Several of these packages some of which I am maintaining offer functionalities for knowledge space theory.

• kst: Basic functionalities for KST (using set representations)
• kstMatrix: Basic functionalities for KST (using more efficient matrix representations)
• kstIO: Functions for loading/saving KST files.

Further R packages for KST include DAKS and pks.

Recently, I have also written an overview on R packages for KST.

R Shiny Apps for Knowledge Space Theory

WIthin the Erasmus+ projects TquanT and QHELP, several R Shiny apps have been developed illustrating and teaching concepts from knowledge space theory.