Filtering Real-Time Linked Data Streams
Name
Mikk-Erik Bachmann
Abstract
The amount of linked data in the Web has increased rapidly in recent years. Linked data, often encoded in RDF, is considered as five-star data in the context of open data due to its usability and potential. Although there has been progress in development of linked data technologies and data processing models, still the full potential of linked data has not been realized. One of the challenges is reasoning over linked data streams, which has just recently gained momentum in research. As a result query languages, such as C-SPARQL, have been proposed and corresponding stream reasoning engines have been implemented. However, such implementations have been evaluated so far mostly in academic settings. This work describes a fully functional proof of concept implementation of a stream reasoning system for message-oriented systems, which is capable of exposing a message queue as a linked data stream, which can be filtered by using C-SPARQL - one of the earliest linked data processing engines. The performance of the C-SPARQL engine, which lies at the heart of the implementation, is evaluated by using CityBench benchmark with settings of an enterprise-scale real-time economy application Inforegister NOW!, which is currently under development.
Graduation Thesis language
English
Graduation Thesis type
Bachelor - Information Technology
Supervisor(s)
Peep Küngas
Defence year
2016