This study is twofold: the first one is the study, using NEO-PIR questionnaire to obtain a description of the main features of the normalpersonality based on a sample of 121 Hispanic persons living in Toulouse. As high levels of depression were found in this sample, the next step was to proceed with the study of a 30 subjects sample, using three questionnaires: NEO PIR, MICM II and SCID with its interview and preliminary observation, focused to found the correlation between personality disturbance and depressive symptoms. The second purpose was to run a test with LAMDA artificial intelligence tool to classify and analyze the results of the questionnaires. This tool is based on fuzzy logic. A quick description of LAMDA algorithm is presented here, making reference to other technical publications. Several modes, learning, self-learning or active and passive recognition are applied in such a way that clinical psychological inferences can be shown.