Simulated hyperlinks on a mobile device
Systems and methods simulate a hyperlink in regular content displayed on a screen. An example method can include generating, responsive to detecting a simulated hyperlink indication, a centered selection from content displayed on a display of a computing device, providing the centered selection to a simulated hyperlink model that predicts an operation given the centered selection, and initiating the operation using an intent associated with a mobile application. The simulated hyperlink model may also provide, from the centered selection, an intelligent selection used the intent's parameter. Another method includes identifying documents having a hyperlink whitelisted websites, generating positive training examples for a simulated hyperlink model using the documents, each positive training example having a centered selection, a web site, and a mobile application mapped to the website, and training the simulated hyperlink model, using the positive training examples, to predict an operation for the mobile application given the centered selection.
1. A method performed on a device, the method comprising:
receiving, at the device and from a server, a simulated hyperlink model, the simulated hyperlink model being a trained neural network model;
using, by the device, training examples generated based on usage of the device to generate updated weights for the simulated hyperlink model, the usage of the device being by a user of the device;
using, by the device, the updated weights to generate an updated simulated hyperlink model personal to the user of the device; and
subsequent to generating the updated simulated hyperlink model and responsive to detecting a tap of the user on a touchscreen display of the device:
automatically generating, based on a location of the tap, an initial selection from text displayed on the touchscreen display the initial selection including a word nearest the location for the tap;
processing the initial selection using the updated simulated hyperlink model to generate output, the updated simulated hyperlink model being stored locally on the device, and the output generated using the updated simulated hyperlink model indicating:
an application installed on the device, and
a refined selection from the text displayed on the touchscreen, the refined selection including the word nearest the location for the tap, and the refined selection differing from the initial selection;
rendering highlighting over the text that corresponds to the refined selection;
rendering a graphical user interface that includes a selectable graphical element that identifies the application;
receiving a selection of the graphical element that identifies the application; and
in response to receiving the selection of the graphical element:
initiating the application with the refined selection as a parameter to cause the application to open in a state that is tailored to the refined selection.
2. The method of claim 1 , wherein initiating the application with the refined selection as the parameter comprises launching the application to a screen that includes content that is based on the refined selection.
3. The method of claim 1 , wherein the refined selection includes a prior word that is prior to the word nearest the location for the tap.
4. The method of claim 3 , wherein the refined selection includes a following word that follows the word nearest the location for the tap.
5. The method of claim 1 ,
wherein using, by the device, the training examples to generate updated weights for the simulated hyperlink model results in a personalized simulated hyperlink model.
6. The method of claim 1 , wherein the selectable graphical element that identifies the application is an icon for the application.
7. The method of claim 6 , wherein the graphical interface further comprises an additional selectable graphical element that, when selected, copies the refined selection to a clipboard.
8. A device comprising:
a touchscreen display;
at least one processor;
a simulated hyperlink model; and
memory storing instructions that, when executed by the at least one processor, cause the mobile device to:
receive, from a server, a simulated hyperlink model, the simulated hyperlink model being a trained neural network model;
use training examples generated based on usage of the device to generate updated weights for the simulated hyperlink model, the usage of the device being by a user of the device;
use the updated weights to generate an updated simulated hyperlink model personal to the user of the device; and
subsequent to generating the updated simulated hyperlink model and responsive to detecting a tap of the user on the touchscreen display of the device:
automatically generate, based on a location of the tap, an initial selection from text displayed on the touchscreen display, the initial selection including a word nearest the location for the tap;
process the initial selection using the updated simulated hyperlink model to generate output, the updated simulated hyperlink model being stored locally on the device, and the output generated using the updated simulated hyperlink model indicating:
an application installed on the device, and
a refined selection from the text displayed on the touchscreen, the refined selection including the word nearest the location for the tap, and the refined selection differing from the initial selection;
render, on the touchscreen display, highlighting over the text that corresponds to the refined selection;
render, on the touchscreen display, a graphical user interface that includes a selectable graphical element that identifies the application;
receive, via the touchscreen display, a selection of the graphical element that identifies the application; and
in response to receiving the selection of the graphical element:
initiate the application with the refined selection as a parameter to cause the application to open in a state that is tailored to the refined selection.
9. The device of claim 8 , wherein instructions that initiate the application with the refined selection as the parameter comprise instructions to launch the application to a screen that includes content that is based on the refined selection.
10. The device of claim 8 , wherein the refined selection includes a prior word that is prior to the word nearest the location for the tap.
11. The device of claim 8 , wherein the refined selection includes a following word that follows the word nearest the location for the tap.
12. The device of claim 8 , wherein the refined selection includes a prior word that is prior to the word nearest the location for the tap and a following word that follows the word nearest the location for the tap.
13. The device of claim 8 , wherein
using, by the device, the training examples to generate updated weights for the simulated hyperlink model results in a personalized simulated hyperlink model.
14. The device of claim 8 , wherein the selectable graphical element that identifies the application is an icon for the application.
15. The device of claim 8 , wherein the graphical interface further comprises an additional selectable graphical element that, when selected, copies the refined selection to a clipboard.