This publication describes techniques and methods that a computing device uses to provide improved query services (e.g., autofill suggestions, speech biasing for automatic speech recognition) to applications on the computing device. To this end, an information collector on the computing device collects application activity information, information displayed on a display, and event information. This collected information can be provided as input to a machine-learned model implemented on the computing device. Responsive to the input received, the machine-learned model can classify the collected information to determine relevant attributes (e.g., keywords, searched locations, names) and make suggestions for utilization by query services provided by the computing device. Through these techniques and methods, user privacy is maintained, less power is consumed by the computing device, and the resources of the computing device (e.g., memory) are conserved.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Shanmugam, Perumaal; Leme, Felipe; and Odell, Julian, "Providing Real-Time Captured Information Input to a Machine-Learned Model to Improve Query Services", Technical Disclosure Commons, (October 29, 2019)