Abstract:

Organizational changes lead Italian financial institutions to review their HR (Human Resources) management model. Therefore, banks need to reorganize themselves in order to rationalize and automate their processes, involving both HR and Training department. A significant starting point for this development is the implementation of an innovative method to map the skills, which should be also capable of giving a complex and clear “snapshot” of the Human Capital, so to preserve its complexity and its dynamic nature.
Currently Italian banks, which do not have an educational mission, analyse skills with methods and instruments which are developed in different institutions and domains of knowledge.
These tools have some particular characteristics: they are attributable to paradigms which are often in contrast, they mix “regulatory” (objective) methods with "based on the person" (subjective) methods. Furthermore, they are mostly built according to a linear measurement system and they are not well calibrated since the organizational context is different.
Basing on this approach, banks implement practices which are often ineffective in defining the criteria for the selection, mobility and training processes, since HR manager also have to consider the industrial relationships in their decision-making.

As an alternative to this view, we propose a new perspective both on the skills and the method to analyse them. According to this theoretical framework, skills are a) nonlinear, b) reticular and c) dynamic entities. More specifically, we define them as: a) "emergent properties" of a series of observable behaviors, which b) are placed and connected among them, according to a distributive logic c) in a defined time period.
Moreover, we believe there are two more levels of “explanatory emergency”: 1) any professional profile is the non-linear combination of multiple skills; 2) human capital is the result of a topological distribution of cognitive-behavioral assets which are associated with professional profiles.
In order to manage this complexity, we have developed a software tool allowing us to represent human capital in a two-dimensional space within which the different features are placed: these elements occupy a specific area depending on the “intensity” levels of the distinctive skills of their role.

We have tested this hypothesis and the tool in analysing the Italian Banking System, using our software to generate "smart pictures" of the human capital considered as a whole: behavioral cognitive assets were placed inside different areas of the map, depending on their properties.
The software integrates a Self-Organizing Map of Kohonen – an Artificial Neural Network - trained by the administration of the FBA Fund data describing each professional-type profile, the so-called "best performers".

In our opinion, the tool has two potential and interesting applications:
1. It could support the HR department during the whole recruitment process: designing training plans, reorganizing educational activities and certificating new professional profiles;
2. It could process the data resulting from the assessment procedures currently used, focusing on the set of knowledge/ability to locate the position of a candidate within the map.
Of course, the model can be extended to other industrial and organizational contexts, according to the specific job profile (knowledge, abilities and skills) related to different professional backgrounds.

Keywords:

Competences, skills, SOM, neural network, software, bank, training, HR.