Context-dependent output volume for voice-controlled virtual assistant