Abstract

Many challenges are being faced when attempting to perform meaningful Social Network Analysis (SNA) on covert networks for intelligence purposes. First, data about covert networks are, by definition, difficult to obtain. Information about those networks is well guarded and, in general, not directly accessible. Consequently, intelligence analysts must build their situational awareness based on an overabundance of indirect information and sources which lead to cluttered heterogeneous models of social networks. This challenge actually results in a second concern in SNA, the imperative to manage very large graphs, which leads to the need to sample or select subsets of the overall data set. Finally, in current systems, analyses of social networks seem to be conducted regardless of the intelligence issue being faced or the data context. This facet is critical in order to ensure that the data lying beneath the analysis are actually truly indicators of the intelligence issue being tackled. This paper first describes the SNA capability targeted along with its challenges. Subsequently, explanations and rationales are provided to highlight the critical roles played by ontologies with respect to the challenges described above. In the current prototype, ontologies are being used with respect to four essential aspects of the SNA capability: to automatically identify and extract the social network data of interest; to organize and correlate these social network data based on the context, to create a filter in order to prune only portions of the social network data; and to select appropriate SNA algorithms corresponding to the intelligence issue being faced. Finally, this paper discusses preliminary results from the implementation of these aspects.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call