Big data and rule-based recommendation system in Internet of Things

This paper proposes a recommendation system based on big data framework and rule-based system in the era of Internet of Things. With the emergence of the smart devices beginning from smart phones extends to the general electronic devices such as smart tv sets, refrigerators, washing machines, robot vacuums, and so on. Such smart devices make it possible to collect the device-usage logs of end users whereby a system is able to analyze it to find the usage patterns of the end users and make recommendations to the end users. Furthermore, this allows to make recommendations on the individual users since the smart devices have their own identifiers such as MAC address and IPv6 address. The smart devices also have matched information with the end user id/s. In this study, we propose a method for analyzing the devise-usage patterns in semi-real time based on the big-data system architecture. We also present a recommendation framework which makes device-usage recommendations by using a rule-based system architecture with the analyzed usage patterns. Lastly, we introduce a segmentation-based analysis and recommendation framework to make recommendations based not only on his or her own usage patterns, but also on the common usage patterns of the users who are living in a similar context. The segmentation is formed also based on the types of the device usages, so that the analysis can be performed in a batch process thereby enabling to make the recommendations in real time based on the pre-analyzed usage patterns.