DATA FUSION APPROACH FOR THE VARIETAL CLASSIFICATION OF LAMBRUSCO P.D.O. WINES

Owing to the increasing of the throughput of modern analytical techniques and the needs of getting exhaustive information on the peculiarities of similar samples in terms of composition for both authenticity and geographical or varietal origin purposes, it is necessary to extract hidden information from complex data set. The possibility to jointly analyze data sets arising from different sources (e.g. different analytical determinations/platforms) allows capturing the latent information that would not be extracted by the individual analysis of each block of data. Several approaches are proposed in the literature and are generally referred to as data fusion ones. In this work a hierarchical data fusion is proposed [1,2] for the characterization of three varieties of Lambrusco Wine (Salamino di Santa Croce, Grasparossa di Castelvetro, Sorbara), a typical P.D.O. product of the District of Modena (Italy). In particular, 60 wine samples of the three different varieties were analyzed by means of HPLC-DAD fro the phenolic compounds evaluation, Emission- Excitation Fluorescence Spectroscopy and 1H-NMR. Since the analytical outputs are characterized by different dimensionality (matrix and tensor), several multivariate analysis were applied (PCA, PARAFAC, MCR-ALS) in order to extract and merge, in a hierarchical way, the information present in each data set. Results showed that this approach was able to well characterize Lambrusco samples giving also the possibility to understand the correlation between the source of information arising from the three analytical techniques.