Towards Publish/Subscribe Functionality on Graphs
Lefteris Zervakis, Christos Tryfonopoulos, Vinay Setty, Stephan Seufert, Spiros Skiadopoulos · Max Planck Digital Library · 2016
In this work, we introduce the publish/subscribe paradigm to support continuous query processing over evolving graphs and motivate it for a number of applications and a variety of possible continuous queries.To the best of our knowledge, this is the first work in the literature that considers supporting publish/subscribe in graphs; we focus specifically on massive and dynamically evolving graphs due to the nature of the problem and the type of targeted applications.To this end, we design a proof-of-concept filtering algorithm for supporting structural matching of continuous graph queries against updates in the evolving graph and demonstrate the need for efficient filtering by experimentally comparing our algorithm against a baseline approach.