MonARCh: an actor based architecture for dynamic linked data monitoring

Area of Science:

• Computer Science
• Data Management
• Web Technologies

Background:

• Applications interacting with the Web of Data require monitoring data sources for changes.
• Existing systems lack effective mechanisms for notifying consumers about data modifications in Linked Data environments.

Purpose of the Study:

• To propose MonARCh, an architecture for monitoring result changes of registered SPARQL queries in Linked Data.
• To provide a robust communication infrastructure for change notification between data sources and consumers.

Main Methods:

• MonARCh functions as a publish/subscribe system, converting user-subscribed SPARQL queries into federated queries.
• It periodically re-executes SERVICE clauses to detect updates and notifies subscribed users.
• The architecture leverages the actor model for concurrent computation and a parallel join algorithm to enhance scalability and query execution speed.

Main Results:

• MonARCh successfully monitors SPARQL query results in Linked Data environments.
• The system demonstrates scalability, capable of monitoring 300 to 25,000 queries within a two-node cluster, depending on query selectivity.
• It meets established requirements for linked data monitoring and incorporates advanced features.

Conclusions:

• MonARCh offers an effective solution for monitoring changes in Linked Data, addressing limitations of existing approaches.
• The architecture's scalability and efficiency make it suitable for real-world applications requiring up-to-date data.
• Design science methodology guided the development and validation of MonARCh, ensuring its practical utility and performance.