This learning support system estimates the user's personal characteristics, hesitation, and degree of concentration from biometric data, such as the handwriting of digital pens, the pressure applied on the pen, eye movement, and status of the pupils, providing appropriate hints.
We proposed plans for this learning support system based on "solving a maze", and developed its UI design, machine learning models, and the app for the exhibition.
"Connected Ink" is an event to promote the development and expansion of digital ink and digital stationery, which Wacom has been holding with DSC since 2016.
Based on the experimental data collected in advance, this machine learning model estimates and classifies user characteristics from the way the maze was solved. Then, depending on the situation, it outputs an optimal hint from a variation of 82 vocals and 7 visual effects.
For example, when an "impatient person" gets lost, it will promptly gives a clear hint like, "Go further up." If a "laid-back person" gets lost, it will leave the person to think and give hints only if the person goes the wrong way.
sdtech continues to research and develop adaptive HMI that will fit to the "state" of the "user."
