Swarm-based approach to evaluate fuzzy classification of semantic sensor data

Sensor networks currently are employed to collect large amounts of heterogeneous data in different and wide environments. Nevertheless, the rapid development and deployment of sensor technology stress the problem related to the availability of too much data and not enough knowledge. Last trend emphasizes the semantic annotation of sensor data. Semantic sensor data increase interoperability between heterogeneous sensor networks and provide contextual information to support situation awareness and management in several application domains. This work defines a framework aimed to reason on distributed semantic sensor data. In particular, defined approach combines swarm intelligence and Fuzzy Control theory in order to infer emerging situation by performing fuzzy classification of semantic sensor data. Swarm architecture enables us to monitor environment by using spatially distributed autonomous sensors. Fuzzy Control theory allows managing of the uncertainty of data sensing. An application scenario for broadcasting traffic news has been described.