Provided are techniques for unified cognition for a virtual personal cognitive assistant. Internet of Things (IoT) devices are coupled to a cognitive model, wherein the cognitive model includes a cognitive classifier, and wherein the cognitive classifier includes a cognitive dimension map and a recognition process. Input from one or more of the IoT devices is received. The cognitive dimension map is used to identify rules based on the input. The recognition process is used to identify events based on the rules. Then, the events are issued to one or more of the IoT devices, wherein the one or more IoT devices execute actions in response to the events.