My focus is user/listener reactions in HCI. Specifically, I want to understand how the strands of the user's nonverbal feedback individually and in combination communicate user comprehension of the agent's vague instructions during an assembly task.
A better learning scheme with dictionary search and with temporal assignment for individual dictionary may help. A huge statistical derivation for user entries will be required as well as suitable learning algorithm which needs a fit for it will be an asset.
- Activity has to be always at human side. So a user is able to learn and communicate as he likes.
- If a user is involved in a learning system ( the computer side has control) the user needs a training of this system before.
- The main goal of HCI is a personal inter-human dialog. Only in this form we have best results of learning because we can use all our senses and biological instruments (hands, feet).
- In HCI we use a computer as instrument between stored information or human dialog partner. So we should use the computer without thinking about the usage.
- HCI should try to use all human senses (seeing, hearing, speaking, touching and not yet possible smelling).
- If we want to get genuine new knowledge we should take care on the possibility of repeating of learning process in every time longer intervals.
So we should see a computer for this reasons as instrument to get an as much as possible direct inter-human communication. That brings most information, motivation and joy.
So the direct answer to your question is: nonverbal communication is a specialised form of HCI and doesn't use all human abilities. It may be soon without personal motivation if the human agent can't steer the dialog.
I form of learning of prewritten texts (perhaps with included explaining links) brings positive motivation - like a legacy book.
Multimodal software may be interesting especially for the goal: getting control of what you have learned already and repeating your knowledge by comparing.