The multitude of curricula and competency models poses great challenges for primary and secondary teachers due to the wealth of descriptions. Defining optimal (or personalized) learning paths is thus impeded. This paper now takes a closer look at 7 curricula from 6 different countries and presents an approach for the identification of learning outcomes and dependencies (requires and expands) between them in order to support the identification of learning paths. The approach includes different strategies from natural language processing, but it also makes use of a refined and simplified version of Bloom’s Taxonomy to identify dependencies between the learning outcomes. It is shown that the identification of similar learning outcomes works very well compared to expert opinions. The identification of dependencies, however, only works well for detecting learning outcomes that refine other learning outcomes (expands dependency). The detection of learning outcomes which build on each other (requires dependency) is, on the other hand, still heavily dependent on the definition of dictionaries and a computing science topics ontology.