When responding to a request from another person, human beings take into account non-verbal cues and relevant implicit contextual information to determine an appropriate response. In contrast, digital assistants do not currently take such contextual information into account. Thus, a digital assistant may provide a sub-optimal or contextually inappropriate response to the user’s request, or require the user to modify the original request iteratively until the user’s full intent is communicated. If the user permits, the techniques described in this disclosure enable a digital assistant to infer the user’s intent by using device and environmental sensors to extract available contextual information, e.g., noise level, the user’s tone of voice, the number of people in a room, objects in the room, temperature, time, location, weather, traffic, etc. The information is used, as permitted by the user, to determine an appropriate response to the user’s request.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.