mintEMU is a text mining project aimed at describing the legacy of the recently discontinued European post-Master of Urbanism (EMU) at TU Delft, Netherlands. The project uses EMU theses published in the TU Delft education repository between the start of the program in 2005 and its closure in 2021. An LDA topic model was used with a number of clusters defined qualitatively according to the thematic set-up of the EMU program. I will discuss the topics, the distribution of theses over topics, and the evolution of topics over the years of the EMU program.

While describing project findings, I will outline the text mining process carried out in R and the development of `mintEMU`, a reproducible research compendium, i.e., an R package comprising a paper, a FAIR dataset and a Shiny dashboard. Building on general-purpose text mining R packages, `mintEMU` adds functionalities needed for the specific structure and content of theses. Based on the experience of the `mintEMU` research compendium, I will discuss the prospect of the general thesis mining package `mint`.

`mintEMU` was developed by Claudiu Forgaci and Aleksandra Wilczynska and can be found at https://github.com/UD3-Lab/mintEMU