Duplication Avoided Data Fusion Technique for Internet of Things

Abstract

Growing technologies and popularities increased number of technology users which tends to generate more volume of data. Handling these data will be more difficult in practical which needs to be fine tuned before giving it for any processing. This is performed in our previous research work by introducing the method namely Semantic based Hierarchical Data Fusion Technique (SHDFT). However computational time of the previous work will be more due to presence of more duplicate data. This is resolved in this research work by introducing the method namely Duplication avoided Data Fusion Technique (DDFT). In this research work, before data fusion, duplication avoidance is done in order to reduce the computational burden of handling repeated data’s. In this research work, initially context based feature extraction is done for extracting the useful features and repeated features. After feature extraction, duplication avoidance is done with the concern of similar concept. After duplication avoidance data fusion is done using hierarchical data fusion technique as done in previous work SHDFT.  Finally performance of the data fusion technique is tested using Improved Convolutional neural network. The overall evaluation of the research work is done in the matlab against existing work in terms of accuracy, precision, recall and f-measure.