A Survey on Knowledge Extraction from WSN

Recently, data management and processing for wireless sensor networks are become a topic of active research in several fields of computer science such as distributed system, database systems and data mining. WSN generates large amount of data streams. Extracting information from the data is the most important task in WSN. Many methods are available to extract the information from WSN. These methods generate the association rules using frequency patterns. It creates large number of rules most of which are non informative and fail to reflect true correlation among the sensor data. This paper presents new type of behavioral pattern called as an associated sensor pattern. Associated sensor pattern captures the co-occurrences and temporal correlation of the sensor data. To capture such patterns a compact tree structure. , called associated sensor pattern tree (ASP-tree) and mining algorithm are proposed. When data streams flows through the network there is possibility of loss of significance for the current time of the old information. ASP tree is further enhanced to SWASP-tree to capture significance of recent data by adopting sliding observation window and updating tree structure accordingly.