Journals Proceedings

Abstract

This position paper advocates applying the monitoring pogwer of IoT to build profiles of user behaviour using the large volumes of collected data. The desired system exploits sensor data mining approaches to profile user behaviour patterns in smart environments. Sensor data is mined to extract relationships of interest between environmental variables (context) and the user, building in this way behaviour profiles. The capability of applying knowledge to manipulate user’s environment is expected take monitoring beyond the simple alert-mode of operation to long term profiling of user’s behaviour. After a brief literature review to prove the suitability of IoT as a low-cost unsupervised profiling platform, we give the details of our proposal and the objectives that needs to be met before user behaviour profiling across inter-spaces is possible.