The methods of big data fusion and semantic collision detection in Internet of Thing

We sometimes find ourselves with plenty of data fusion in Internet of Thing, which necessitates an automatic removing semantic collision. For this, it is necessary to detect semantic collision, with a fairly reliable method to find many semantic collision and powerful enough to run in a reasonable time. Big data fusion in Internet of Thing represents today an important data quality challenge which leads to bad decision-making. This paper proposes and compares on real data effective fusion matching methods for automatic removing semantic collision of files based on names, working with Chinese texts or English texts, and the names of people or places, in East or in the West. After conducting a more complete classification of big data fusion than the usual classifications, we introduce several methods for big data fusion. Through a simple model, we highlight a global efficiency, accuracy and recover. We propose a new measuring mechanism between records, as well as rules for automatic big data fusion. Analyses made on Internet of Thing containing real data in western cities, and on a known standard Internet of Thing containing names of companies in the China, have shown better results than those of known methods, with a lesser complexity.