A semantic reasoner using attributed graphs based on intelligent fusion of security multi-sources information

Abstract

Recently, the need of monitoring both real and virtual environments is growing up, especially in security contexts. Virtual environments are rich of data produced by human interactions that can not be extracted using classical physical sensors. Thus, new kind of sensors allow to obtain and collect a huge quantity of data from these virtual environment. In order to monitor complex environments, in which the human factor is essential, arises the need of combining both data derived from objective measurements (hard data) and data derived from human interaction (soft data). In this paper we present a method and a software architecture for the fusion of heterogeneous data. The novelty of this method is the joint use of a rule-based inference engine, of a graph matcher and of semantic ontology reasoning to combine and process structured data coming for hard and soft sources. An application of the proposed system is presented within the framework of a Security Intelligence project.