Smart Connection of User Profiles in a Big Data Context

Abstract

The idea we propose in this article is a follow-up to our research series in the ontology-based profiling framework. The approach relies on tracking user profile changes for user connection within a Big Data context. We have worked in our series of research on the identification and qualification of profiles in web 2.0 context based on the ontological approach and multi agent system. Among the limitations of our research is the fact that changing interests over time does not affect the relationships between profiles. The goal of our approach is to follow the change of the interests of internet users and to propose afterwards new relations having changed activities in the same direction. In order to implement this approach, we will first use the ontology approach. The ontology approach will allows describing semantic models and determine the properties, restrictions and axioms of our application domain. The ontology we propose will generate a set of domains and sub domains of activities used to identify user profiles. On the other hand, we will use the Multi Agents approach to process users' activities before classifying them in their profiles.