Inferring activities of an actor by means of context ontologies

The present article describes an architecture proposed to identify the probable actions taken by an actor involved in a software development process. Such a process is a collaborative activity that can be associated to a context that comprises tasks and interactions for information exchange, targeting the manipulation of artifacts being developed. In this work, context is described using ontologies with concepts related to activities, events and devices. The events initiated by the actor are detected trough sensors, e. g. data about the execution platform, IDE data collected with the Hackystat [1] tool, web navigation data etc. Finally a scenario is introduced to illustrate the application of the proposed architecture.