This disclosure describes techniques for automatic generation and presentation of communication parameters via a virtual assistant dialog to a user that is engaged in another activity, e.g., driving a vehicle. The parameters are generated locally based on user-permitted contextual factors. Users can select from the presented options by using any combination of voice input, touch/gesture input, etc. The generated parameters can include, e.g., call/message recipients, message content, etc. The recipients are suggested based on prior on-device activity. Upon user selection of a message recipient, message content options are provided based on ranking candidate messages based on the context. With user permission, appropriate contextual information, e.g.. a destination, current location, etc. is included in the body of the candidate message.

Creative Commons License

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