Automation of web services invocations on a mobile device

The article describes the task of automating web services invocations on a mobile device in dependence to the current context. We have defined the statement of the problem of classifying context for invoking web services. It has been revealed that for this task Wi-Fi access points are the most accessible information on a mobile device inside the building. We have developed a method for classifying the contexts for web services application on a mobile device, which implements classification based on learning with a teacher. In the process of learning, the user indicates positive and negative cases for invoking services. To assess the use cases for existing Wi-Fi access points, we have introduced a metric that takes into account the signal strength and the possible lack of information from them. The classifier is built for each combination of service and user. The results of the study confirm the efficiency of the applying method and the possibility of its use in smart home and smart city systems.