In this contribution, we apply data-driven techniques to investigate fundamental characteristics of dislocation-based plasticity. We analyze the kinetics of dislocation interactions of complex dislocation networks in three-dimensional discrete dislocation dynamics data. By using first principle techniques of data mining, we introduce a dislocation density-based rate formulation for the dislocation network evolution on each slip system. As a key feature, the formulation includes reaction kinetic parameters for each type of dislocation reaction in face-centred cubic (fcc) systems. We apply our formulation to tensile tested fcc aluminum single-crystals. A central observation is the dependency of the reaction kinetics with respect to the crystal orientation and slip system activity. The results are discussed for a comparison between discrete and continuum modeling. Finally, we particularly focus on the discussion of a meaningful application of data-driven techniques to investigate fundamental microstructural characteristics in dislocation-based plasticity and material homogenization.