(2)In recent years, the techniques of credit risk assessment and the calculation of credit risk scores, in addition to more traditional statistical techniques, have been developed under innovative approaches based on artificial intelligence and machine learning. For example, for small and medium-sized enterprises without a long credit history, innovative methods based on transactional data may prove more useful than methods based on traditional balance sheet data. Due to their complexity, the use of those techniques may increase the asymmetry of information between investors and crowdfunding service providers. Therefore, the description of the method used by crowdfunding service providers to calculate credit scores should indicate the scoring model used to support such calculation, as well as sufficient information on the financial and non-financial factors that are used as input in those scoring models and on the output provided by the scoring models.