Summary
The easy deployment of wireless sensors allows the development of context‐aware applications that could react to the environment changes and users preferences. For example, data extracted from mobile phones, IoT devices, and embedded computers in buses and taxis could be processed in real time to understand city dynamics. However, gathering and real‐time processing of relevant data are still a challenging task. For instance, the high volume of streaming data requires real‐time processing in order to generate immediate mitigation actions. Complex‐event processing (CEP) techniques and predictive analytics have been recently proposed for analyzing streaming data in order to generate fast insights and then take suitable actions according to the situations' changes. These techniques could be also used to predict future situations and react to them properly before happening. The work presented in this paper focuses mainly on the performance evaluation of three CEP engines, CQELS, C‐SPARQL, and ETALIS, that are widely used by researchers for linked stream data processing. Experiments have been conducted using two existing benchmarks, CityBench and SP2Bench. Several performance metrics have been evaluated to assess on their ability to process high streaming data using complex queries. Results are reported to show the efficiency and scalability of these CEP engines for both social‐based data (SP2Bench) and physical‐based data (CityBench). Reported results show that ETALIS outperforms CQELS and C‐SPARQL in terms of throughput and memory utilization. ETALIS was integrated as a use case scenario for occupancy and air quality comfort in energy‐efficient buildings. Results show the usefulness of CEP for real‐time data processing and reasoning.