Ontological approach for competence management using K-NN with R

Abstract

With the advent of the movement of the immaterial economy which leads us towards an economy and a company of knowledge, the importance is given to the concepts of knowledge and competencies for better immaterial capital development. Consequently, the management of the „human capital‟ became both a necessary and a complex matter for the Human Resources professionals who are concerned about employee‟s intellectual capitals (knowledge management) all along with their competences. The management of competence is a topic that has always interested in several research teams and was the object of several publications of scientific work. The research of qualified staff is one of the most important axes of this work. It lies in the scope of the decision-making aid for the selection of one person starting from a group of candidates in order to achieve the goals of employment with well defined characteristics. In this context, we present various approaches referring to a competence, to its definitions, its characterizations and its methods of treatment. Our first contribution will be directed towards the proposal of a typology and a characterization of a job and competences and a system of evaluation of the level of these competences. The semantic web technologies and those of the data mining are chosen for the implementation of this contribution. Indeed, competence is a field of knowledge with excellence. Thus, the field of the semantic web offers panoply of standards to represent knowledge and to handle them. The whole of this knowledge will be represented in the form of ontology that will be treated with the external resource. WordNet, in order to guide the research of candidates with the qualified skills to occupy a given job. The k-nearest neighbor algorithm, or quite simply kNN and the languages R and SPARQL are used for this implementation.