Abstract:
The rapid and significant increase in the amount of sensor data to be processed requires the use of techniques to reduce the size of data in order to efficiently extract the relevant knowledge. In this paper, we present two approaches used to derive data summaries. The first one relies on linguistic quantifiers in the sense of Yager. The second one leverages the notion of the typical value of a data set. Then, we show our implementation of these two methods with some experiments conducted on different databases (real-flight data collected from the ADSB project and real data for smart city collected from neOCampus project). The comparative studies show the best approach w.r.t. execution time.