Data mining in IoT: data analysis for a new paradigm on the internet

This paper provides an overview on Data Mining (DM) technologies for the Internet of Things (IoT). IoT has become an active area of research, since IoT promises among other to improve quality of live and safety in Smart Cities, to make resource supply and waste management more efficient, and optimize traffic. DM is highly domain specific and depends on what is being mined for. For instance, if IoT is used to optimize traffic in a Smart City to reduce traffic jams and to find parking spaces quicker, different types of data needs to be collected and analysed from an eHealth solution, where IoT is used in a Smart Home to monitor the well being of patients or elderly people. IoT connects things that can collect numeric data from smart sensors, streaming data from cameras or route information on maps. Depending on the type of data, different techniques need to be adopted to analyse them. Also, many IoT applications analyse data from different devices and correlate them to make predictions about possible machine failures in production sites or looming emergency situations in Smart Buildings in a home security application. DM techniques need to handle the heterogeneity of IoT data, the large volumes of data and the speed at which they are produced. This paper explores the state of the art DM techniques for IoT.