Because the Internet is evolving to host increasingly complex content and online applications in the future, it is no longer effective to manage such content and applications manually. Previous researchers have introduced layers for automated network management of today’s networks. Autonomic layers use a publish-subscribe model to help autonomic elements communicate with each other. The autonomic layers use filters to accomplish subscription. However, the filters are statically defined by the consumers. Such statically defined subscription filters limit the scalability of context dissemination in large-scale networks.

The authors focus on designing an automated system for generating subscription filters using a semantic context model. They first define an ontological context model and then illustrate how data can be stored in the context model. Based on the semantic context model, the authors present a context translation algorithm that translates the contextual requirements of a rule-based system to the context model. This section also describes two detailed algorithms that automatically generate subscription filters based on policies from network operators using OWL and SWRL or RDF and SPARQL. The performance evaluation shows some trade-off between these two approaches and proves that the proposed approach works and is scalable to many autonomic elements.

The significant contribution of this paper is in generating automated subscription filters inside an administrative domain. A few previous works have focused on this topic, but none conducted detailed research on trade-offs among different subscription filter generation algorithms. By evaluating different generation algorithms, the authors show the scalability of their approach, which is significant.