Building a user profile for GUI preferences.
Publication Date: 2017-Aug-04
The IP.com Prior Art Database
Building a User Profile for GUI Preferences
Adaptive user interfaces are important in order to display information in user-friendly fashion. A system is here described that can infer user preferences "live" based on light conditions and estimate of user fatigue. This disclosure describes a system capable of building a user profile to store their preferences, as well as estimating default user preferences if no information is available for current user.
While a user interacts with a device (computer monitor, smartphone, tablet, TV, virtual/augmented reality headset), the GUI setting on said device is adjusted automatically to reduce user strain and maximise reading speed and comfort under current ambient conditions.
Ambient conditions include light and distance from device. GUI settings includes font type, size, contrast and brightness. Once a setting is found that is deemed good for current user under current ambient condition said setting is stored as part of the user’s profile.
The next time said user engages with a device (not necessarily the same device), the GUI settings learnt earlier will be available and presented as a default to the user.
User sensor comprises of: - any means available to estimate user biometrics, including e.g. cameras
for face recognition, keyboard for keystroke dynamics and any combination of the above.
- any means available to estimate user condition Cu, including e.g. cameras for expression recognition and eye tracking, microphone to estimate stress level in voice and any combination of the above.
Ambient sensor comprises: - any means available to estimate significant ambient conditions Ca that can
affect user visual perception, including e.g. light sensor to estimate illumination, camera to estimate user distance from screen.
User profile can be seen as a map the given current (Ca, Cu) provide a value P(Ca, Cu) for user preferences under current conditions.
As a user engages with a device, a recognition procedure is carried out. If a match is identified for current user (i.e. returning user for whom a profile
already exists), user profile is loaded and GUI is displayed based on current measurements from ambient sensor and user sensor. As a heuristic, this implies measuring ambient condition Ca, user condition Cu, retrieving from user profile the preference P(Ca, Cu). For instance, "With low illumination at a distance of 60 cm from display (ambient condition), if user displays no sign of tiredness (user condition) display text in serif 12 with medium brightness".
If no match is identified (new user), exploit...