Abstract
Mobile devices and wearable computers can detect a variety of events to automatically trigger actions. However, it is hard to determine the level of user satisfaction with such automatically triggered actions. This disclosure describes techniques, implemented with user permission, to determine a user’s reaction to automatically triggered actions. Upon the execution of an automatically triggered action, sensors on a wearable device are utilized to measure physiological parameters such as changes in heart rate, blood oxygen concentration, blood pressure, etc., and/or mechanical movements. Physiological and movement data are used to automatically generate ground-truth labels for a triggered action based on the resulting level of user satisfaction. The labeled actions can be used to update machine learning models that trigger automatic actions.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
n/a, "Use of Data from Physiological Sensors for Automatic Labeling of User Sentiment", Technical Disclosure Commons, (July 07, 2023)
https://www.tdcommons.org/dpubs_series/6033