Presenting an approach to optimizing the storage and search of spatiotemporal information in a semantic format, this paper makes a fine contribution to the semantic web and sensor data processing communities.
The purpose of the proposed method is to improve the efficiency of searches for information coming from sensor data. An open-source semantic spatiotemporal data engine (SSTDE) combines standard geographic information system (GIS) tools with semantic tools in a hybrid architecture, constructing a subgraph index instead of the usual single-node indices, with a query optimization algorithm based on the subgraph index. The advantages of the proposed approach are in the hybrid architecture; the query optimization; and the composite index, which indexes subgraphs, and thus enables search through a graph-matching algorithm. An evaluation benchmark for testing the approach consists of five queries with different complex graph patterns, which are run on five systems independent from the triple store used to store and query the data. The experiments show that SSTDE performs considerably better than the other four approaches (OpenRDF, Neo4j, Parliament, and Neo4j) in terms of speed of response.
I found this to be a technically well-motivated paper with a sufficient level of detail and a clear explanation of the motivation and advantages of the chosen approach. The paper will be worth reading for engineers and software developers interested in building solutions for big data and the semantic web, including but not limited to spatiotemporal and sensor data.