Exploiting Context Information in Spoken Dialogue Interaction with Mobile Devices ?