A systematic and automated comparison between the physical quantities reconstructed from experimental data and the ones resulted from physical models could be a very useful approach for long pulse tokamak experiments producing a large amount of data. A discrepancy in such comparison brings key information for the scientific exploitation of the experiment since it means either that the pulse deviates from usual models i.e. there is interesting physics to analyze in that pulse or that there are issues with the reconstructed data, coming from the measurements or from the way they have been processed, which have to be corrected. In this paper we propose to develop an expert system that carries out in an integrated way i) a reconstruction of plasma quantities from the measurements, ii) a prediction of the reconstructed quantities, according to validated expectations / models, iii) and an intelligent comparison of the first two steps. The paper shows the first application of this quite general comparison method to the electron density profile reconstruction in the core plasma.