TY  - JOUR
T1  - Tolerance Based Adaptable Framework for Proficient Massive Data Mining in Wireless Sensor Networks
AU - Nagara, J. AU - Nawaz, G.M. Kadhar 
JO  - Asian Journal of Information Technology
VL  - 12
IS  - 6
SP  - 190
EP  - 197
PY  - 2013
DA  - 2001/08/19
SN  - 1682-3915
DO  - ajit.2013.190.197
UR  - https://makhillpublications.co/view-article.php?doi=ajit.2013.190.197
KW  - Data mining
KW  -wireless sensor network
KW  -fault tolerance
KW  -neural network
KW  -bayesian data sharing
KW  -k-nearest neighbor
AB  - Wireless Sensor Network (WSN) is more popular because of its wide application in real-time. Sensor network provides immensely valuable information from the environment where it is implemented. Consequently, effective data mining is required in sensor networks. Processing huge data set is difficult in WSN due to limited power, memory space and short-range communication. Various data mining algorithms for WSN has been proposed but all existing algorithms fails to handle massive data set. To process huge data set effectively researchers constructed a framework in this study. Moreover, structure forms a cluster in a highly coverage area of WSN among the nodes. Cluster formation is carried through neural networks. It uses Bayesian data sharing technique to monitor the activity of database and data sharing. Along with communication problems, mobility of nodes causes break in communication and also leads to fault occurrence. In order to face the mobility issues, researchers proposed a fault tolerance method through creating an optimal sensor network. The experimental result demonstrates the framework is efficient in handling huge data set.
ER  -