[en] The aim of the work presented in this talk is to measure the atmospheric electron neutrino spectrum on data collected by the IceCube Detector. IceCube is a Cherenkov neutrino detector instrumenting 1 km${}^3$ of ice at the South Pole. The Cherenkov light detected by IceCube is emitted by charged particles like atmospheric muons or the secondary particles of neutrino interactions passing through the detector. Events caused by neutrinos result in different event topologies, depending on the neutrino flavor and the kind of interaction. The events this analysis is interested in, the interaction of electron neutrinos, have a spherical topology in the detector. These events are referred to as cascade-like events. Muons, either atmospheric or from muon neutrino interactions, show an elongated event topology and are referred to as track-like events. The first part of this analysis is to build an event sample containing a high amount of cascade-like events with low amounts of track-like events, utilizing machine learning methods. Subsequently an unfolding approach will be used to obtain to the atmospheric electron neutrino spectrum. This analysis is currently planned to encompass one year of data from the year 2012, however a extension to multiple years may be possible in the near future. This talk will give an overview over the current status of this analysis and its prospects.