Context aware and pattern oriented machine learning framework(CAPOMF) for Android

Abstract

Context-aware computing refers to a very peculiar kind of application that can sense the physical environment and behave accordingly. This feature of context awareness can be further enhanced using machine learning. This is possible by providing services not only related to the usage pattern of the users but also the environmental context of the user. Android being one of the most popular mobile platform has constantly addressed context awareness through hardware and software support. This paper evaluates the Android platform and also proposes a framework that incorporates context-aware services along with machine learning capability in order to provide a one of a kind pattern oriented framework that not only aids in development of applications but also utilizes the hardware support of sensors of the Android hardware system to utilize the full potential of the mobile platform. Finally this paper also evaluates the framework by building an application that helps to solve the transit problems arising due to potholes present in the streets of Mumbai. The application built effectively maps the locations of such potholes and helps commuters avoid such potholes and thus saves the repair costs of vehicles.