An Integrating Architecture for Competence Management

Software applications dealing with human resources and their skills, attitudes, and knowledge (such as e-learning systems, skills databases, e-recruitment portals, corporate portals integrated with competence centered services, and functionalities) are often based on database technology (usually relational) for storing, organizing, and searching relevant information. These existing applications have two major limitations. First, they and their databases are based on raw data (such as CVs and job offers and job/role descriptions), which are organized according to some ad-hoc “reference grid” (like a job or trade tree): indeed, limited attention is devoted to data organization and to its foundations. Data organization should be based on the central concept of competency: raw data are interesting if they convey information about what abilities are required for accomplishing tasks and what abilities individuals hold (or have acquired); this information is indeed forming the competence, required and acquired respectively. Second, applications based on database technology do not really support the systematic analysis, exploration, and sharing of raw data and therefore offer limited support and weak integration to what can be called competence management processes. For instance, within a process for assessing individual competencies, it is difficult to implement portal services that try to automatically find out competencies of individuals from their CVs or, inside a company, from other documents (like activity or process reports which individuals have made). Unfortunately, despite a huge amount of work, there is no consensus on the competency definition. This is especially because most of the current work prioritizes some processes over other processes (e.g., evaluating competencies is prioritized over identifying needed competencies). Worse, as usual, some works prioritize enabling technologies over models. This results in partial or overloaded models for representing what competency, acquired and required, is; additionally, there is no clear picture of limitations of these models because prioritized processes are often hidden