Abstract:
In this work, we apply the semantic information push para- digm in the domain of cultural heritage and advocate for its usefulness in a number of diverse scenarios ranging from personalised content delivery at museum visitors to the curation of huge knowledge bases that integrate diverse cultural assets. We envision a large-scale semantic information push system that is able to perform efficient filtering of multiple RDF data streams based on expressive subscriptions that aim both for the structure and content of the stream. To this end, we put forward STIP, a new algorithm that indexes user subscriptions and utilises its index structures to efficiently match them against the stream of RDF events; STIP proves four orders of magnitude faster than its baseline competitor in an experimental evaluation with real-world data. To the best of our knowledge, this is the first approach in the literature to propose the usage of information push –along with an appropriate algorithm– as the technological substrate for a variety of high-level cultural heritage applications such as personalisation and recommender systems.