Time Series Distributed Analysis in IoT with ETL and Data Mining Technologies

Abstract

The paper describes an approach to performing a distributed analysis on time series. The approach suggests to integrate Data Mining and ETL technologies and to perform primary analysis of time series based on a subset of data sources (primary data sources). Other data sources are only used if it is necessary to obtain additional information. This allows to reduce the number of requests to data sources and network traffic. In the result it makes it possible to use communication channels with low bandwidth (including wireless networks) for data collection.