U.S. Department of Energy

Pacific Northwest National Laboratory

Remembering the Important Things: Semantic Importance in Stream Reasoning

Publish Date: 
Tuesday, October 18, 2016
Reasoning and querying over data streams rely on the ability to deliver a sequence of stream snapshots to the processing algorithms. These snapshots are typically provided using windows as views into streams and associated window management strategies. Generally, the goal of any window management strategy is to preserve the most important data in the current window and preferentially evict the rest, so that the retained data can continue to be exploited. A simple timestamp- based strategy is rst-in-rst-out (FIFO), in which items are replaced in strict order of arrival. All timestamp-based strategies implicitly assume that a temporal ordering reliably reects importance to the processing task at hand, and thus that window management using timestamps will maximize the ability of the processing algorithms to deliver accurate interpretations of the stream. In this work, we explore a general notion of semantic importance that can be used for window management for streams of RDF data using semantically-aware processing algorithms like deduction or semantic query. Semantic importance exploits the information carried in RDF and surrounding ontologies for ranking window data in terms of its likely contribution to the processing algorithms. We explore the general semantic categories of query contribution, provenance, and trustworthiness, as well as the contribution of domain-specific ontologies. We describe how these categories behave using several concrete examples. Finally, we consider how a stream window management strategy based on semantic importance could improve overall processing performance, especially as available window sizes decrease.
Yan R, MT Greaves, WP Smith, and DL McGuinness. 2016. "Remembering the Important Things: Semantic Importance in Stream Reasoning." In Stream Reasoning Workshop 2016, October 18, 2016, Kobe, Japan.
| Pacific Northwest National Laboratory