Abstract: Method for controlling data quality assurance after a migration between a source repository (1), comprising source data associated with a source data structure, and a target repository, comprising target data associated with a target data structure, according to migration rules, wherein said target data is organized as a set of elements; comprising: simulating (S1) said migration by applying migration rules to source data so as to output simulated data according to said target data structure, said simulated data being organized as a set of elements; merging (S4) the simulated data and the target data by excluding identical elements, to form an input vector for a machine learning algorithm; applying (S5) this machine learning algorithm for pairing elements of said input vector according to a similarity score; determining (S6) discrepancies indicator according to said similarity score for each element of the input vector.