Keywords: Behavioral big data, emotions, self-selection treatment, multidimensional data
The work intends to propose the solution of some methodological issues of treatment of behavioural big data, with particular regard to data science applications for business. The issue of the construction of semi-automatic auto-selection adjustment systems is particularly important in applications that produce data-driven decision-making systems. Micro-marketing applications are an important example of this, and humanization of some large database will be discussed in the work through the application of indirect psychometric models such as Semiometry and Limbic analysis.