The edge platform has evolved to become a part of a distributed computing environment. While typical edges do not have enough processing power to train machine learning models in real time, it is common to generate models in the cloud for use on the edge. The pattern of heterogeneous Internet of Things (IoT) data is dependent on individual circumstances. It is not easy to guarantee prediction performance when a monolithic model is used without considering the spatial characteristics of the space generating those data. In this paper, we propose a collaborative framework using a new method to select the best model for the edge from candidate models of cloud based on sample data correlation. This method lets the edge use the most suitable model without any training tasks on the edge side, and it also minimizes privacy issues. We apply the proposed method to predict future fine particulate matter concentration in an individual space. The results suggest that our method can provide better performance than the previous method.

No abstract

No abstract

: Various personalized services are provided based on user context these days, and IoT(Internet of Things) devices provides effective ways to collect user context. For example, user's activity such as walking steps, calories, and sleeping hours can be collected using smart activity tracker. Smart scale can sense change of user's weight or body fat percentage. However, these services are independent to each other and not easy to make them collaborate. Many standard bodies are working on the documents for this issue, but due to diversity of IoT use case scenarios, it seems that multiple IoT technologies co-exist for the time being. This paper propose a framework to collaborate heterogeneous IoT services. The proposed framework provides methods to build application for heterogeneous IoT devices and user context management in more intuitive way using HTTP. To improve compatibility and usability, gathered user contexts are based on MPEG-UD. Implementation of framework and service with real-world devices are also presented.

No abstract