Seppo Törmä

Processing Heterogeneous RDF Events with Standing SPARQL Update Rules

SPARQL query language is targeted to search datasets encoded in RDF. SPARQL Update adds support of insert and delete operations between graph stores, enabling queries to process data in steps, have persistent memory and communicate with each other. When used in a system supporting incremental evaluation of multiple simultaneously active and collaborating queries SPARQL can define entire event processing networks. The method is demonstrated by an example service, which triggers notifications about the proximity of friends, comparing alternative SPARQL-based approaches. Observed performance in terms of both notification delay and correctness of results far exceed systems based on window repetition without extending standard SPARQL or RDF.

Semantic Linking of Building Information Models

Information is produced and maintained in a building project in a concurrent and loosely coupled manner by a set of parties representing different disciplines. The multiple partial models produced in a project are interrelated since they represent different aspects of a same building. The standard conceptual schema of buildings - Industry Foundation Classes (IFC) - enables conceptual, type-level compatibility between models but solves interoperability problems only partially. We argue for the need for instance-level interoperability. The capability to link the individuals of different models together would support functionalities such as cross-model information access and aggregation, status monitoring, and change management. The nature of cross-model linking is investigated using a small case study and a set of principles for the implementation of linking are identified. The natural way to implement the principles using Linked Data technologies is outlined. In a qualitative evaluation against previous approaches, the linking approach turns out to be more flexible and expressive but more demanding of support from design tools and information management infrastructure.

Efficient matching of SPARQL subscriptions using rete

Ubiquitous domains such as smart spaces, location-aware mobile systems, or internet-of-things are characterized by large and volatile sets of heterogeneous and independently behaving entities like devices, services, and other identified objects. This study focuses on efficient implementation of an event processing system to manage interaction among these entities. The approach is based on expressive semantic representations: information sharing in RDF and content-based publish/subscribe with SPARQL as the subscription language. SPARQL can be used to construct elaborate queries for detecting complex states resulting from receiving events produced by multiple interrelated entities. The notification system should aim at short notification times while simultaneously allowing high throughput of events. We study incremental matching of SPARQL queries on RDF data using Rete algorithm. The results obtained demonstrate that an efficient and fast semantic notification framework can be implemented by representing SPARQL queries and RDF triples as rules and facts in a Rete engine.