Generating predictive action buttons within a graphical user interface
Examples described herein include systems for generating a predictive action button within a graphical user interface. The system can launch an action generator that tracks a user's behavior, either directly or through the use of an action profile. User behavior can be tracked in individual applications or across multiple applications and devices. Action modules can implement application events and procedures to track particular events and behavior information for that application. A module engine can analyze correlations between the events and behaviors. Based on this analysis, the predictive action button can be displayed for carrying out the predicted action when the corresponding application event occurs.
1. A system for generating a predictive action button within a graphical user interface, the system comprising:
a non-transitory, computer-readable medium that contains instructions;
a processor that executes the instructions to perform stages including:
tracking a user's behavior with respect to at least two applications, wherein each of the two applications executes on a different physical device associated with the user;
aggregating the tracked behavior from the at least two applications at an action profile manager executing on a first of the different physical devices;
reformatting the tracked behavior by the action profile manager according to instructions specified by an action profile stored at the first of the different physical devices;
transmitting the reformatted tracked behavior from the first of the different physical devices to a server remote from the different physical devices;
based on the reformatted tracked behavior with respect to the at least two applications, identifying a probability of at least one potential action of the user;
analyzing the reformatted tracked behavior based on the user's interactions with the different physical devices;
determining, based on the analyzing, a likelihood of the at least one potential action of the user on each of the physical devices;
based on the probability meeting a threshold value, generating and displaying the predictive action button on a subset of the physical devices based on the determined likelihood of the at least one potential action exceeding a likelihood threshold, wherein the predictive action button relates to the at least one potential action of the user; and
carrying out the at least one potential action based on the user selecting the generated predictive action button.
2. The system of claim 1 , wherein tracking the user's behavior further comprises monitoring the user's behavior based on the action profile for the user.
3. The system of claim 2 , wherein identifying a probability of at least one potential action of the user comprises calculating said probability based, at least partially, on the action profile.
4. The system of claim 1 , wherein the predictive action button is displayed within an email application, proximate a particular email, and wherein carrying out the at least one potential action comprises causing the email application to delete only the particular email.
5. The system of claim 1 , wherein identifying the probability is performed at a server that is remote from the graphical user interface.
6. The system of claim 1 , wherein tracking and identifying the probability are performed by a single application.
7. The system of claim 1 , wherein identifying the probability is based on a historical average time for performing the action of the user.
8. A non-transitory, computer-readable medium containing instructions that, when executed by a processor, performs stages for generating a predictive action button within a graphical user interface, the stages comprising:
tracking a user's behavior with respect to at least two applications, wherein each of the two applications executes on a different physical device associated with the user;
aggregating the tracked behavior from the at least two applications at an action profile manager executing on a first of the different physical devices;
reformatting the tracked behavior by the action profile manager according to instructions specified by an action profile stored at the first of the different physical devices;
transmitting the reformatted tracked behavior from the first of the different physical devices to a server remote from the different physical devices;
based on the reformatted tracked behavior with respect to the at least two applications, identifying a probability of at least one potential action of the user;
analyzing the reformatted tracked behavior based on the user's interactions with the different physical devices;
determining, based on the analyzing, a likelihood of the at least one potential action of the user on each of the physical devices;
based on the probability meeting a threshold value, generating and displaying the predictive action button on a subset of the physical devices based on the determined likelihood of the at least one potential action exceeding a likelihood threshold, wherein the predictive action button relates to the at least one potential action of the user; and
carrying out the at least one potential action based on the user selecting the generated predictive action button.
9. The non-transitory, computer-readable medium of claim 8 , wherein tracking the user's behavior further comprises monitoring the user's behavior based on the action profile for the user.
10. The non-transitory, computer-readable medium of claim 9 , wherein identifying a probability of at least one potential action of the user comprises calculating said probability based, at least partially, on the action profile.
11. The non-transitory, computer-readable medium of claim 8 , wherein the predictive action button is displayed within an email application, proximate a particular email, and wherein carrying out the at least one potential action comprises causing the email application to delete only the particular email.
12. The non-transitory, computer-readable medium of claim 8 , wherein identifying the probability is performed at a server that is remote from the graphical user interface.
13. The non-transitory, computer-readable medium of claim 8 , wherein tracking and identifying the probability are performed by a single application.
14. The non-transitory, computer-readable medium of claim 8 , wherein identifying the probability is based on a historical average time for performing the action of the user.
15. A method for generating a predictive action button within a graphical user interface, comprising:
tracking a user's behavior with respect to at least two applications, wherein each of the two applications executes on a different physical device associated with the user;
aggregating the tracked behavior from the at least two applications at an action profile manager executing on a first of the different physical devices;
reformatting the tracked behavior by the action profile manager according to instructions specified by an action profile stored at the first of the different physical devices;
transmitting the reformatted tracked behavior from the first of the different physical devices to a server remote from the different physical devices;
based on the reformatted tracked behavior with respect to the at least two applications, identifying a probability of at least one potential action of the user;
analyzing the reformatted tracked behavior based on the user's interactions with the different physical devices;
determining, based on the analyzing, a likelihood of the at least one potential action of the user on each of the physical devices;
based on the probability meeting a threshold value, generating and displaying the predictive action button on a subset of the physical devices based on the determined likelihood of the at least one potential action exceeding a likelihood threshold, wherein the predictive action button relates to the at least one potential action of the user; and
carrying out the at least one potential action based on the user selecting the generated predictive action button.
16. The method of claim 15 , wherein tracking the user's behavior further comprises monitoring the user's behavior based on the action profile for the user.
17. The method of claim 16 , wherein identifying a probability of at least one potential action of the user comprises calculating said probability based, at least partially, on the action profile.
18. The method of claim 15 , wherein the predictive action button is displayed within an email application, proximate a particular email, and wherein carrying out the at least one potential action comprises causing the email application to delete only the particular email.
19. The method of claim 15 , wherein identifying the probability is performed at a server that is remote from the graphical user interface.
20. The method of claim 15 , wherein tracking and identifying the probability are performed by a single application.